Classifif=cation ameliore.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "id": "b0dd9286",
      "metadata": {
        "id": "b0dd9286"
      },
      "source": [
        "-----------------------------------------------------------------------------------------------------------------------\n",
        "**-- Membres du groupe** :\n",
        ">   * Soufiane Eghazi 13°\n",
        ">   * Ibtissam Labyady 23°\n",
        ">   * Amine Maasri  26°\n",
        "\n",
        "**-- fillière** : *ICSD*\n",
        "##\n",
        "---\n",
        "\n",
        "## Projet machine learning\n",
        "\n",
        "---\n",
        "-----------------------------------------------------------------------------------------------------------------------"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "05e55ed1",
      "metadata": {
        "id": "05e55ed1"
      },
      "source": [
        "#### 5.Entraîner différents modèles : \n",
        "    On va créer différents modèle de machine learning afin de comparer leurs performances aprés ,\n",
        "    -----------------------------------------------------------------------------------------------------------------------"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "6a3150b6",
      "metadata": {
        "id": "6a3150b6"
      },
      "source": [
        "###### 5.1 Classification:\n",
        "    On va commencer par les méthodes de ML supérvise de classification \n",
        "    -----------------------------------------------------------------------------------------------------------------------"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "aa42cab2",
      "metadata": {
        "id": "aa42cab2"
      },
      "source": [
        "##### 5.1.1 With feature selection 1:\n",
        "    avec les variables choisit:\n",
        "    -----------------------------------------------------------------------------------------------------------------------"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "f9187b00",
      "metadata": {
        "id": "f9187b00"
      },
      "outputs": [],
      "source": [
        "#KNN\n",
        "\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "model_1_with = KNeighborsClassifier(n_neighbors=1)\n",
        "model_1_with.fit(X_train_selected, Y_train)\n",
        "Y_predec_model_1_with =model_1_with.predict(X_test_selected)\n",
        "Y_predec_model_1_with_train =model_1_with.predict(X_train_selected)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "e2b15ae4",
      "metadata": {
        "id": "e2b15ae4"
      },
      "outputs": [],
      "source": [
        "# Arbre de décision\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "model_2_with = DecisionTreeClassifier(max_depth=7, random_state=42)\n",
        "model_2_with.fit(X_train_selected, Y_train)\n",
        "Y_predec_model_2_with = model_2_with.predict(X_test_selected)\n",
        "Y_predec_model_2_with_train = model_2_with.predict(X_train_selected)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "ccc32199",
      "metadata": {
        "id": "ccc32199",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f877b56f-abdc-44a1-9707-bde399643f9f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/sklearn/linear_model/_logistic.py:458: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
            "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
            "\n",
            "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
            "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
            "Please also refer to the documentation for alternative solver options:\n",
            "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
            "  n_iter_i = _check_optimize_result(\n"
          ]
        }
      ],
      "source": [
        "# Régression logistique\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "model_3_with = LogisticRegression(random_state=0, C=10)\n",
        "model_3_with.fit(X_train_selected, Y_train)\n",
        "\n",
        "Y_predec_model_3_with = model_3_with.predict(X_test_selected)\n",
        "Y_predec_model_3_with_train = model_3_with.predict(X_train_selected)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "28ac1392",
      "metadata": {
        "id": "28ac1392"
      },
      "outputs": [],
      "source": [
        "# Forêt aléatoire\n",
        "\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "model_4_with = RandomForestClassifier(n_estimators=115, random_state=0)\n",
        "model_4_with.fit(X_train_selected, Y_train)\n",
        "Y_predec_model_4_with =model_4_with.predict(X_test_selected)\n",
        "Y_predec_model_4_with_train =model_4_with.predict(X_train_selected)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "7e00285b",
      "metadata": {
        "id": "7e00285b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b4194b68-2a56-4cd6-a135-8ee089b997a9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/sklearn/svm/_base.py:1244: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
            "  warnings.warn(\n"
          ]
        }
      ],
      "source": [
        "# SVM\n",
        "\n",
        "from sklearn.svm import LinearSVC\n",
        "model_5_with = LinearSVC(random_state=0, tol=1e-5,C=1)\n",
        "model_5_with.fit(X_train_selected, Y_train)\n",
        "Y_predec_model_5_with =model_5_with.predict(X_test_selected)\n",
        "Y_predec_model_5_with_train =model_5_with.predict(X_train_selected)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "d328dcb4",
      "metadata": {
        "id": "d328dcb4"
      },
      "outputs": [],
      "source": [
        "# Naive Bayes\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "model_6_with = GaussianNB(var_smoothing=1e-09)\n",
        "model_6_with.fit(X_train_selected, Y_train)\n",
        "Y_predec_model_6_with =model_6_with.predict(X_test_selected)\n",
        "Y_predec_model_6_with_train =model_6_with.predict(X_train_selected)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import cross_val_score\n",
        "\n",
        "models = {\n",
        "    'KNN': KNeighborsClassifier(n_neighbors=1),\n",
        "    'Decision Tree': DecisionTreeClassifier(random_state=42,max_depth76),\n",
        "    'Logistic Regression': LogisticRegression(random_state=0,C=10),\n",
        "    'Random Forest': RandomForestClassifier(n_estimators=115, random_state=0),\n",
        "    'SVM': LinearSVC(random_state=0, tol=1e-5,C=1),\n",
        "    'Naive Bayes': GaussianNB(var_smoothing=1e-09)\n",
        "}\n",
        "\n",
        "scoring = 'accuracy' # choose your desired scoring metric\n",
        "\n",
        "results = {}\n",
        "for name, model in models.items():\n",
        "    scores = cross_val_score(model, X_all_train, Y_all_train, cv=5, scoring=scoring)\n",
        "    results[name] = scores.mean()\n",
        "\n",
        "best_model = max(results, key=results.get)\n",
        "print('Results:', results)\n",
        "print('Best model:', best_model)\n"
      ],
      "metadata": {
        "id": "jWAHiu-tFJWh"
      },
      "id": "jWAHiu-tFJWh",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "RandomForestClassifier(n_estimators=115, random_state=0)"
      ],
      "metadata": {
        "id": "_HD0OBDkpKcS"
      },
      "id": "_HD0OBDkpKcS"
    },
    {
      "cell_type": "markdown",
      "id": "9dc36499",
      "metadata": {
        "id": "9dc36499"
      },
      "source": [
        "##### 5.1.2 With feature selection 2:\n",
        "    avec les deux composantes APC\n",
        "    -------------------------------------------------------------------------------------------------------------------------"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "d8f62e74",
      "metadata": {
        "id": "d8f62e74"
      },
      "outputs": [],
      "source": [
        "#KNN\n",
        "\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "model_1_withacp = KNeighborsClassifier(n_neighbors=10)\n",
        "model_1_withacp.fit(X_train_pca, Y_train)\n",
        "Y_predec_model_1_withacp =model_1_withacp.predict(X_test_pca)\n",
        "Y_predec_model_1_withacp_train =model_1_withacp.predict(X_train_pca)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "aae601dc",
      "metadata": {
        "id": "aae601dc"
      },
      "outputs": [],
      "source": [
        "# Arbre de décision\n",
        "\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "model_2_withacp = DecisionTreeClassifier(random_state=42,max_depth=6)\n",
        "model_2_withacp.fit(X_train_pca, Y_train)\n",
        "Y_predec_model_2_withacp =model_2_withacp.predict(X_test_pca)\n",
        "Y_predec_model_2_withacp_train =model_2_withacp.predict(X_train_pca)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "01be7df7",
      "metadata": {
        "id": "01be7df7"
      },
      "outputs": [],
      "source": [
        "# Régression logistique\n",
        "\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "model_3_withacp = LogisticRegression(random_state=0,C=1)\n",
        "model_3_withacp.fit(X_train_pca, Y_train)\n",
        "Y_predec_model_3_withacp =model_3_withacp.predict(X_test_pca)\n",
        "Y_predec_model_3_withacp_train =model_3_withacp.predict(X_train_pca)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "db32e42a",
      "metadata": {
        "id": "db32e42a"
      },
      "outputs": [],
      "source": [
        "# Forêt aléatoire\n",
        "\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "model_4_withacp = RandomForestClassifier(n_estimators=74, random_state=0)\n",
        "model_4_withacp.fit(X_train_pca, Y_train)\n",
        "Y_predec_model_4_withacp =model_4_withacp.predict(X_test_pca)\n",
        "Y_predec_model_4_withacp_train =model_4_withacp.predict(X_train_pca)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "1b70d68e",
      "metadata": {
        "id": "1b70d68e",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ab7d34e8-82dc-4aea-ba8f-f99dd0a42ec2"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/sklearn/svm/_base.py:1244: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
            "  warnings.warn(\n"
          ]
        }
      ],
      "source": [
        "# SVM\n",
        "\n",
        "from sklearn.svm import LinearSVC\n",
        "model_5_withacp = LinearSVC(random_state=0, tol=1e-5,C=1)\n",
        "model_5_withacp.fit(X_train_pca, Y_train)\n",
        "Y_predec_model_5_withacp =model_5_withacp.predict(X_test_pca)\n",
        "Y_predec_model_5_withacp_train =model_5_withacp.predict(X_train_pca)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "d8375d09",
      "metadata": {
        "id": "d8375d09"
      },
      "outputs": [],
      "source": [
        "# Naive Bayes\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "model_6_withacp = GaussianNB(var_smoothing= 1e-09)\n",
        "model_6_withacp.fit(X_train_pca, Y_train)\n",
        "Y_predec_model_6_withacp =model_6_withacp.predict(X_test_pca)\n",
        "Y_predec_model_6_withacp_train =model_6_withacp.predict(X_train_pca)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import cross_val_score\n",
        "\n",
        "models = {\n",
        "    'KNN': KNeighborsClassifier(n_neighbors=10),\n",
        "    'Decision Tree': DecisionTreeClassifier(random_state=42,max_depth=6),\n",
        "    'Logistic Regression': LogisticRegression(random_state=0,C=1),\n",
        "    'Random Forest': RandomForestClassifier(n_estimators=74, random_state=0),\n",
        "    'SVM': LinearSVC(random_state=0, tol=1e-5,C=1),\n",
        "    'Naive Bayes': GaussianNB(var_smoothing=1e-09)\n",
        "}\n",
        "\n",
        "scoring = 'accuracy' # choose your desired scoring metric\n",
        "\n",
        "results = {}\n",
        "for name, model in models.items():\n",
        "    scores = cross_val_score(model, X_all_train, Y_all_train, cv=5, scoring=scoring)\n",
        "    results[name] = scores.mean()\n",
        "\n",
        "best_model = max(results, key=results.get)\n",
        "print('Results:', results)\n",
        "print('Best model:', best_model)\n"
      ],
      "metadata": {
        "id": "mwP-E6HoEzdZ"
      },
      "id": "mwP-E6HoEzdZ",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "Zd5zsrpbEipb"
      },
      "id": "Zd5zsrpbEipb",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "RandomForestClassifier(n_estimators=74, random_state=0),\n",
        "    "
      ],
      "metadata": {
        "id": "2BUScEu3puSc"
      },
      "id": "2BUScEu3puSc"
    },
    {
      "cell_type": "markdown",
      "id": "a0b0b60e",
      "metadata": {
        "id": "a0b0b60e"
      },
      "source": [
        "##### 5.1.3 Without feature selection:\n",
        "    avec toutes les variables \n",
        "    -------------------------------------------------------------------------------------------------------------------------"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "0b51ad98",
      "metadata": {
        "id": "0b51ad98"
      },
      "outputs": [],
      "source": [
        "#KNN\n",
        "\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "model_1_without = KNeighborsClassifier(n_neighbors=1)\n",
        "model_1_without.fit(X_all_train, Y_all_train)\n",
        "Y_predec_model_1_without =model_1_without.predict(X_all_test)\n",
        "Y_predec_model_1_without_train =model_1_without.predict(X_all_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "7f0881ca",
      "metadata": {
        "id": "7f0881ca"
      },
      "outputs": [],
      "source": [
        "# Arbre de décision\n",
        "\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "model_2_without = DecisionTreeClassifier(random_state=42,max_depth=10)\n",
        "model_2_without.fit(X_all_train, Y_all_train)\n",
        "Y_predec_model_2_without =model_2_without.predict(X_all_test)\n",
        "Y_predec_model_2_without_train =model_2_without.predict(X_all_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "a6f77712",
      "metadata": {
        "id": "a6f77712",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "460b74e1-6925-48fe-97b1-76458f46fc7f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/sklearn/linear_model/_logistic.py:458: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
            "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
            "\n",
            "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
            "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
            "Please also refer to the documentation for alternative solver options:\n",
            "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
            "  n_iter_i = _check_optimize_result(\n"
          ]
        }
      ],
      "source": [
        "# Régression logistique\n",
        "\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "model_3_without = LogisticRegression(random_state=0,C=10)\n",
        "model_3_without.fit(X_all_train, Y_all_train)\n",
        "Y_predec_model_3_without =model_3_without.predict(X_all_test)\n",
        "Y_predec_model_3_without_train =model_3_without.predict(X_all_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "b6abcfb9",
      "metadata": {
        "id": "b6abcfb9"
      },
      "outputs": [],
      "source": [
        "# Forêt aléatoire\n",
        "\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "model_4_without = RandomForestClassifier(n_estimators=91, random_state=0)\n",
        "model_4_without.fit(X_all_train, Y_all_train)\n",
        "Y_predec_model_4_without =model_4_without.predict(X_all_test)\n",
        "Y_predec_model_4_without_train =model_4_without.predict(X_all_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "bda520a2",
      "metadata": {
        "id": "bda520a2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "50b646a7-b78d-4f03-a688-8b6c464f990a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/sklearn/svm/_base.py:1244: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
            "  warnings.warn(\n"
          ]
        }
      ],
      "source": [
        "# SVM\n",
        "\n",
        "from sklearn.svm import LinearSVC\n",
        "model_5_without = LinearSVC(random_state=0, tol=1e-5,C=1)\n",
        "model_5_without.fit(X_all_train, Y_all_train)\n",
        "Y_predec_model_5_without =model_5_without.predict(X_all_test)\n",
        "Y_predec_model_5_without_train =model_5_without.predict(X_all_train)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "873e34a1",
      "metadata": {
        "id": "873e34a1"
      },
      "outputs": [],
      "source": [
        "# Naive Bayes\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "model_6_without = GaussianNB(var_smoothing=1e-05)\n",
        "model_6_without.fit(X_all_train, Y_all_train)\n",
        "Y_predec_model_6_without =model_6_without.predict(X_all_test)\n",
        "Y_predec_model_6_without_train =model_6_without.predict(X_all_train)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import cross_val_score\n",
        "\n",
        "models = {\n",
        "    'KNN': KNeighborsClassifier(n_neighbors=1),\n",
        "    'Decision Tree': DecisionTreeClassifier(random_state=42,max_depth=10),\n",
        "    'Logistic Regression': LogisticRegression(random_state=0,C=10),\n",
        "    'Random Forest': RandomForestClassifier(n_estimators=91, random_state=0),\n",
        "    'lg':LinearSVC(random_state=0, tol=1e-5,C=1),\n",
        "    'Naive Bayes': GaussianNB(var_smoothing=1e-05)\n",
        "}\n",
        "\n",
        "scoring = 'accuracy' # choose your desired scoring metric\n",
        "\n",
        "results = {}\n",
        "for name, model in models.items():\n",
        "    scores = cross_val_score(model, X_all_train, Y_all_train, cv=5, scoring=scoring)\n",
        "    results[name] = scores.mean()\n",
        "\n",
        "best_model = max(results, key=results.get)\n",
        "print('Results:', results)\n",
        "print('Best model:', best_model)"
      ],
      "metadata": {
        "id": "SJwj0mc3Ecby"
      },
      "id": "SJwj0mc3Ecby",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        " RandomForestClassifier(n_estimators=91, random_s"
      ],
      "metadata": {
        "id": "CWpEMeUtERUM"
      },
      "id": "CWpEMeUtERUM"
    },
    {
      "cell_type": "markdown",
      "id": "d87eddc4",
      "metadata": {
        "id": "d87eddc4"
      },
      "source": [
        "## III.Évaluation des performances du modèle :\n",
        "    Après avoir construit des modèles, on va évaluer leurs performances en utilisant des métriques appropriées telles que la précision, le rappel et la F-mesure\n",
        "    -----------------------------------------------------------------------------------------------------------------------"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "3c925870",
      "metadata": {
        "id": "3c925870"
      },
      "source": [
        "#### 1.Classification :"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "4a781d7f",
      "metadata": {
        "id": "4a781d7f"
      },
      "source": [
        "##### Mesures de performances:"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "3bf06bc7",
      "metadata": {
        "id": "3bf06bc7"
      },
      "source": [
        "###### 1.1 With feature selection 1:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "aafc962f",
      "metadata": {
        "id": "aafc962f"
      },
      "outputs": [],
      "source": [
        "def plot_confusion_matrix(y_test,y_pred, cmap):\n",
        "    classes=['class 1', 'class 2','class 3']\n",
        "    # Créer la matrice de confusion\n",
        "    conf_mat = confusion_matrix(y_test, y_pred)\n",
        "\n",
        "    # Afficher la matrice de confusion avec les valeurs formatées\n",
        "    fig, ax = plt.subplots(figsize=(8, 8))\n",
        "    im = ax.imshow(conf_mat, cmap=cmap)\n",
        "\n",
        "    # Ajouter les valeurs dans la matrice de confusion\n",
        "    for i in range(conf_mat.shape[0]):\n",
        "        for j in range(conf_mat.shape[1]):\n",
        "            ax.text(j, i, format(conf_mat[i, j], 'd'),\n",
        "                    ha=\"center\", va=\"center\", color=\"white\")\n",
        "\n",
        "    # Ajouter les noms des classes et du graphe\n",
        "    ax.set_title(\"Matrice de confusion\")\n",
        "    ax.set_xticks(range(conf_mat.shape[1]))\n",
        "    ax.set_xticklabels(classes)\n",
        "    ax.set_yticks(range(conf_mat.shape[0]))\n",
        "    ax.set_yticklabels(classes)\n",
        "    ax.set_ylabel('Vraie classe')\n",
        "    ax.set_xlabel('Classe prédite')\n",
        "    plt.colorbar(im)\n",
        "\n",
        "    # Afficher la figure\n",
        "    plt.show()\n",
        "   \n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "4d0e8eac",
      "metadata": {
        "id": "4d0e8eac",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "880d3b30-629a-4d22-9ded-7ff95e861e75"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___KNeighborsClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       1.00      1.00      1.00      1322\n",
            "     class 2       1.00      1.00      1.00       231\n",
            "     class 3       1.00      1.00      1.00       147\n",
            "\n",
            "    accuracy                           1.00      1700\n",
            "   macro avg       1.00      1.00      1.00      1700\n",
            "weighted avg       1.00      1.00      1.00      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1322    0    0]\n",
            " [   0  231    0]\n",
            " [   0    0  147]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 1.0\n",
            "______________________________Test______________________________________\n",
            "                     ___KNeighborsClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.96      0.98      0.97       333\n",
            "     class 2       0.86      0.77      0.81        64\n",
            "     class 3       0.93      0.93      0.93        29\n",
            "\n",
            "    accuracy                           0.94       426\n",
            "   macro avg       0.92      0.89      0.90       426\n",
            "weighted avg       0.94      0.94      0.94       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[326   6   1]\n",
            " [ 14  49   1]\n",
            " [  0   2  27]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9436619718309859\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import classification_report\n",
        "target_names = ['class 1', 'class 2','class 3']\n",
        "\n",
        "\n",
        "plot_confusion_matrix(Y_test,Y_predec_model_1_with, 'PiYG')   \n",
        "\n",
        "                        ###########################################\n",
        "                        ############___KNeighborsClassifier___#####\n",
        "                        ###########################################\n",
        "print(\"______________________________Train______________________________________\")              \n",
        "print(\"                     ___KNeighborsClassifier___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_1_with_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_1_with_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_KNN_with = accuracy_score(Y_train, Y_predec_model_1_with_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_KNN_with)\n",
        "print(\"______________________________Test______________________________________\")              \n",
        "print(\"                     ___KNeighborsClassifier___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_1_with, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_1_with))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_KNN_with = accuracy_score(Y_test, Y_predec_model_1_with)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_KNN_with)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Comparing the train and test results, we can see that the classifier is overfitting on the training data as the accuracy on the training set is 100%, while the accuracy on the test set is lower at 94.36%. The precision, recall, and f1-score for all three classes are also lower on the test set compared to the training set. This indicates that the model is not generalizing well to unseen data."
      ],
      "metadata": {
        "id": "cCP1pPAVLwE6"
      },
      "id": "cCP1pPAVLwE6"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "bc347d66",
      "metadata": {
        "id": "bc347d66",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "9580653f-edcc-49ee-8928-6d838df69dca"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___DecisionTreeClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.97      0.99      0.98      1322\n",
            "     class 2       0.89      0.85      0.87       231\n",
            "     class 3       1.00      0.95      0.98       147\n",
            "\n",
            "    accuracy                           0.96      1700\n",
            "   macro avg       0.96      0.93      0.94      1700\n",
            "weighted avg       0.96      0.96      0.96      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1303   19    0]\n",
            " [  35  196    0]\n",
            " [   3    4  140]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9641176470588235\n",
            "______________________________Test______________________________________\n",
            "                     ___DecisionTreeClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.96      0.96      0.96       333\n",
            "     class 2       0.83      0.81      0.82        64\n",
            "     class 3       0.90      0.97      0.93        29\n",
            "\n",
            "    accuracy                           0.94       426\n",
            "   macro avg       0.90      0.91      0.90       426\n",
            "weighted avg       0.94      0.94      0.94       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[320  11   2]\n",
            " [ 11  52   1]\n",
            " [  1   0  28]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9389671361502347\n"
          ]
        }
      ],
      "source": [
        "                        ###########################################\n",
        "                        ############___DecisionTreeClassifier___###\n",
        "                        ###########################################\n",
        "            \n",
        "plot_confusion_matrix(Y_test,Y_predec_model_2_with, 'PiYG')            \n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                     ___DecisionTreeClassifier___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_2_with_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_2_with_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_DecisionTree_with = accuracy_score(Y_train, Y_predec_model_2_with_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_DecisionTree_with)\n",
        "print(\"______________________________Test______________________________________\")   \n",
        "print(\"                     ___DecisionTreeClassifier___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_2_with, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_2_with))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_DecisionTree_with = accuracy_score(Y_test, Y_predec_model_2_with)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_DecisionTree_with)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        " Le modèle a une exactitude de 0,938, ce qui signifie que le modèle prédit correctement environ 94 % des données de test."
      ],
      "metadata": {
        "id": "foe2VnfHMo2_"
      },
      "id": "foe2VnfHMo2_"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "ee6401dd",
      "metadata": {
        "id": "ee6401dd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "0340d8bd-9b5f-4b93-ac7a-62933feda707"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___LogisticRegression___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.94      0.95      0.94      1322\n",
            "     class 2       0.64      0.60      0.62       231\n",
            "     class 3       0.79      0.76      0.77       147\n",
            "\n",
            "    accuracy                           0.89      1700\n",
            "   macro avg       0.79      0.77      0.78      1700\n",
            "weighted avg       0.88      0.89      0.88      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1258   51   13]\n",
            " [  76  138   17]\n",
            " [   9   27  111]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8864705882352941\n",
            "______________________________test______________________________________\n",
            "                     ___LogisticRegression___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.93      0.94      0.93       333\n",
            "     class 2       0.65      0.64      0.65        64\n",
            "     class 3       0.75      0.72      0.74        29\n",
            "\n",
            "    accuracy                           0.88       426\n",
            "   macro avg       0.78      0.77      0.77       426\n",
            "weighted avg       0.88      0.88      0.88       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[312  17   4]\n",
            " [ 20  41   3]\n",
            " [  3   5  21]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8779342723004695\n"
          ]
        }
      ],
      "source": [
        "                        ###########################################\n",
        "                        ############___LogisticRegression___#######\n",
        "                        ###########################################\n",
        "        \n",
        "plot_confusion_matrix(Y_test,Y_predec_model_3_with, 'PiYG')   \n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                     ___LogisticRegression___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_3_with_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_3_with_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_LogisticRegression_with = accuracy_score(Y_train, Y_predec_model_3_with_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_LogisticRegression_with)\n",
        "print(\"______________________________test______________________________________\") \n",
        "print(\"                     ___LogisticRegression___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_3_with, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_3_with))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_LogisticRegression_with = accuracy_score(Y_test, Y_predec_model_3_with)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_LogisticRegression_with)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "The two models have similar performance on both training and testing data. The first model has slightly lower precision and recall for class 2 and class 3 in both training and testing data compared to the second model, but the overall accuracy is slightly higher for the first model on the training data. However, the accuracy on the testing data is slightly higher for the second model.\n",
        "\n",
        "In summary, the two models have very similar performance and it is difficult to say which one is better without further analysis or comparison with other models."
      ],
      "metadata": {
        "id": "OLxHib8kmFtH"
      },
      "id": "OLxHib8kmFtH"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "002f38fd",
      "metadata": {
        "id": "002f38fd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "88650960-fc6f-44ca-aa5c-d416313ff179"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___RandomForestClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       1.00      1.00      1.00      1322\n",
            "     class 2       1.00      1.00      1.00       231\n",
            "     class 3       1.00      1.00      1.00       147\n",
            "\n",
            "    accuracy                           1.00      1700\n",
            "   macro avg       1.00      1.00      1.00      1700\n",
            "weighted avg       1.00      1.00      1.00      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1322    0    0]\n",
            " [   0  231    0]\n",
            " [   0    0  147]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 1.0\n",
            "______________________________test______________________________________\n",
            "                     ___RandomForestClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.97      0.98      0.97       333\n",
            "     class 2       0.90      0.83      0.86        64\n",
            "     class 3       0.90      0.93      0.92        29\n",
            "\n",
            "    accuracy                           0.95       426\n",
            "   macro avg       0.92      0.91      0.92       426\n",
            "weighted avg       0.95      0.95      0.95       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[326   5   2]\n",
            " [ 10  53   1]\n",
            " [  1   1  27]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9530516431924883\n"
          ]
        }
      ],
      "source": [
        "                        ###########################################\n",
        "                        ############___RandomForestClassifier___###\n",
        "                        ###########################################\n",
        "\n",
        "\n",
        "plot_confusion_matrix(Y_test,Y_predec_model_4_with, 'PiYG')   \n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                     ___RandomForestClassifier___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_4_with_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_4_with_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_RandomForest_with = accuracy_score(Y_train, Y_predec_model_4_with_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_RandomForest_with)\n",
        "print(\"______________________________test______________________________________\")             \n",
        "print(\"                     ___RandomForestClassifier___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_4_with, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_4_with))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_RandomForest_with = accuracy_score(Y_test, Y_predec_model_4_with)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_RandomForest_with)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "the first has Accuracy: 0.9553990610328639"
      ],
      "metadata": {
        "id": "UYO4yPf3mSrI"
      },
      "id": "UYO4yPf3mSrI"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "be47b503",
      "metadata": {
        "id": "be47b503",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "d3ddc875-ffa4-4683-f7c9-9d2eb9b276ce"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                      ___Support Vector Machine (SVM)___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.93      0.95      0.94      1322\n",
            "     class 2       0.64      0.57      0.60       231\n",
            "     class 3       0.82      0.77      0.79       147\n",
            "\n",
            "    accuracy                           0.89      1700\n",
            "   macro avg       0.80      0.76      0.78      1700\n",
            "weighted avg       0.88      0.89      0.88      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1261   50   11]\n",
            " [  85  132   14]\n",
            " [  10   24  113]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8858823529411765\n",
            "______________________________Test______________________________________\n",
            "               ___Support Vector Machine (SVM)___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.93      0.95      0.94       333\n",
            "     class 2       0.68      0.61      0.64        64\n",
            "     class 3       0.70      0.72      0.71        29\n",
            "\n",
            "    accuracy                           0.88       426\n",
            "   macro avg       0.77      0.76      0.76       426\n",
            "weighted avg       0.88      0.88      0.88       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[315  13   5]\n",
            " [ 21  39   4]\n",
            " [  3   5  21]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8802816901408451\n"
          ]
        }
      ],
      "source": [
        "                        ###########################################\n",
        "                        ####___Support Vector Machine (SVM) ___####\n",
        "                        ###########################################\n",
        "\n",
        "plot_confusion_matrix(Y_test,Y_predec_model_5_with, 'PiYG')    \n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                      ___Support Vector Machine (SVM)___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_5_with_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_5_with_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_SVM_with = accuracy_score(Y_train, Y_predec_model_5_with_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_SVM_with)\n",
        "print(\"______________________________Test______________________________________\")               \n",
        "print(\"               ___Support Vector Machine (SVM)___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_5_with, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_5_with))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_SVM_with = accuracy_score(Y_test, Y_predec_model_5_with)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_SVM_with)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "the same\n"
      ],
      "metadata": {
        "id": "pb8NOT1KmcAP"
      },
      "id": "pb8NOT1KmcAP"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "4fd994f2",
      "metadata": {
        "id": "4fd994f2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "ba901ee0-a3d0-4d3e-bf13-1a3ae6d8e8ce"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                      ______Naives Bayes______\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.97      0.87      0.92      1322\n",
            "     class 2       0.53      0.84      0.65       231\n",
            "     class 3       0.65      0.68      0.66       147\n",
            "\n",
            "    accuracy                           0.85      1700\n",
            "   macro avg       0.72      0.79      0.74      1700\n",
            "weighted avg       0.89      0.85      0.86      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1148  132   42]\n",
            " [  25  193   13]\n",
            " [   5   42  100]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8476470588235294\n",
            "______________________________Test______________________________________\n",
            "               ______Naives Bayes______\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.97      0.83      0.89       333\n",
            "     class 2       0.50      0.80      0.61        64\n",
            "     class 3       0.51      0.69      0.59        29\n",
            "\n",
            "    accuracy                           0.81       426\n",
            "   macro avg       0.66      0.77      0.70       426\n",
            "weighted avg       0.87      0.81      0.83       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[276  44  13]\n",
            " [  7  51   6]\n",
            " [  1   8  20]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8145539906103286\n"
          ]
        }
      ],
      "source": [
        "                        ###########################################\n",
        "                        ##########_______naive bayes _______#######\n",
        "                        ###########################################\n",
        "\n",
        "plot_confusion_matrix(Y_test,Y_predec_model_6_with, 'PiYG')   \n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                      ______Naives Bayes______\")            \n",
        "print(classification_report(Y_train, Y_predec_model_6_with_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_6_with_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_naive_Bayes_with = accuracy_score(Y_train, Y_predec_model_6_with_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_naive_Bayes_with)\n",
        "print(\"______________________________Test______________________________________\")               \n",
        "print(\"               ______Naives Bayes______\")            \n",
        "print(classification_report(Y_test, Y_predec_model_6_with, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_6_with))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_naive_Bayes_with = accuracy_score(Y_test, Y_predec_model_6_with)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_naive_Bayes_with)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "the same"
      ],
      "metadata": {
        "id": "Nnfbb8_tmgy9"
      },
      "id": "Nnfbb8_tmgy9"
    },
    {
      "cell_type": "markdown",
      "id": "9cf3f5ef",
      "metadata": {
        "id": "9cf3f5ef"
      },
      "source": [
        "###### 1.2 With feature selection 2:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "fd0c59df",
      "metadata": {
        "id": "fd0c59df",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "3b586ac9-a49e-45b4-eba4-ad701e1023bf"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___KNeighborsClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.92      0.97      0.94      1322\n",
            "     class 2       0.71      0.58      0.64       231\n",
            "     class 3       0.88      0.67      0.76       147\n",
            "\n",
            "    accuracy                           0.89      1700\n",
            "   macro avg       0.84      0.74      0.78      1700\n",
            "weighted avg       0.89      0.89      0.89      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1282   29   11]\n",
            " [  93  135    3]\n",
            " [  24   25   98]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8911764705882353\n",
            "______________________________Test______________________________________\n",
            "                     ___KNeighborsClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.91      0.97      0.94       333\n",
            "     class 2       0.72      0.52      0.60        64\n",
            "     class 3       0.80      0.69      0.74        29\n",
            "\n",
            "    accuracy                           0.88       426\n",
            "   macro avg       0.81      0.72      0.76       426\n",
            "weighted avg       0.87      0.88      0.87       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[322   8   3]\n",
            " [ 29  33   2]\n",
            " [  4   5  20]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8802816901408451\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_test, Y_predec_model_1_withacp, 'RdBu')  \n",
        "\n",
        "                        ###########################################\n",
        "                        ############___KNeighborsClassifier___#####\n",
        "                        ###########################################\n",
        "print(\"______________________________Train______________________________________\")              \n",
        "print(\"                     ___KNeighborsClassifier___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_1_withacp_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_1_withacp_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_KNN_withacp = accuracy_score(Y_train, Y_predec_model_1_withacp_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_KNN_withacp)\n",
        "print(\"______________________________Test______________________________________\")              \n",
        "print(\"                     ___KNeighborsClassifier___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_1_withacp, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_1_withacp))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_KNN_withacp = accuracy_score(Y_test, Y_predec_model_1_withacp)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_KNN_withacp)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "the first has Accuracy: 0.8994117647058824"
      ],
      "metadata": {
        "id": "AjN7evTdmn-M"
      },
      "id": "AjN7evTdmn-M"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "e3f5930c",
      "metadata": {
        "id": "e3f5930c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "98ae2dd2-eefe-4384-d207-bb0b40b3a8d4"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___DecisionTreeClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.93      0.97      0.95      1322\n",
            "     class 2       0.71      0.64      0.67       231\n",
            "     class 3       0.99      0.73      0.84       147\n",
            "\n",
            "    accuracy                           0.90      1700\n",
            "   macro avg       0.88      0.78      0.82      1700\n",
            "weighted avg       0.90      0.90      0.90      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1282   39    1]\n",
            " [  83  148    0]\n",
            " [  17   22  108]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9047058823529411\n",
            "______________________________Test______________________________________\n",
            "                     ___DecisionTreeClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.89      0.95      0.92       333\n",
            "     class 2       0.64      0.44      0.52        64\n",
            "     class 3       0.80      0.69      0.74        29\n",
            "\n",
            "    accuracy                           0.86       426\n",
            "   macro avg       0.78      0.69      0.73       426\n",
            "weighted avg       0.85      0.86      0.85       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[318  12   3]\n",
            " [ 34  28   2]\n",
            " [  5   4  20]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8591549295774648\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_test, Y_predec_model_2_withacp, 'RdBu')   \n",
        "                        ###########################################\n",
        "                        ############___DecisionTreeClassifier___###\n",
        "                        ###########################################\n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                     ___DecisionTreeClassifier___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_2_withacp_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_2_withacp_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_DecisionTree_withacp = accuracy_score(Y_train, Y_predec_model_2_withacp_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_DecisionTree_withacp)\n",
        "print(\"______________________________Test______________________________________\")   \n",
        "print(\"                     ___DecisionTreeClassifier___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_2_withacp, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_2_withacp))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_DecisionTree_withacp = accuracy_score(Y_test, Y_predec_model_2_withacp)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_DecisionTree_withacp)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "0e4d0640",
      "metadata": {
        "id": "0e4d0640",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "a63263cf-aac7-4e46-a793-7686a0e63a02"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___LogisticRegression___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.89      0.96      0.92      1322\n",
            "     class 2       0.62      0.42      0.50       231\n",
            "     class 3       0.66      0.53      0.59       147\n",
            "\n",
            "    accuracy                           0.85      1700\n",
            "   macro avg       0.72      0.64      0.67      1700\n",
            "weighted avg       0.83      0.85      0.84      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1265   26   31]\n",
            " [ 123   98   10]\n",
            " [  35   34   78]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8476470588235294\n",
            "______________________________test______________________________________\n",
            "                     ___LogisticRegression___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.90      0.96      0.93       333\n",
            "     class 2       0.69      0.45      0.55        64\n",
            "     class 3       0.57      0.55      0.56        29\n",
            "\n",
            "    accuracy                           0.86       426\n",
            "   macro avg       0.72      0.66      0.68       426\n",
            "weighted avg       0.85      0.86      0.85       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[321   6   6]\n",
            " [ 29  29   6]\n",
            " [  6   7  16]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8591549295774648\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_test, Y_predec_model_3_withacp, 'RdBu')   \n",
        "                        ###########################################\n",
        "                        ############___LogisticRegression___#######\n",
        "                        ###########################################\n",
        "        \n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                     ___LogisticRegression___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_3_withacp_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_3_withacp_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_LogisticRegression_withacp = accuracy_score(Y_train, Y_predec_model_3_withacp_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_LogisticRegression_withacp)\n",
        "print(\"______________________________test______________________________________\") \n",
        "print(\"                     ___LogisticRegression___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_3_withacp, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_3_withacp))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_LogisticRegression_withacp = accuracy_score(Y_test, Y_predec_model_3_withacp)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_LogisticRegression_withacp)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "e1b8e8d5",
      "metadata": {
        "id": "e1b8e8d5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "51cf7811-ce8e-4161-96d9-d21cd16fd24c"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___RandomForestClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       1.00      1.00      1.00      1322\n",
            "     class 2       1.00      1.00      1.00       231\n",
            "     class 3       1.00      0.99      1.00       147\n",
            "\n",
            "    accuracy                           1.00      1700\n",
            "   macro avg       1.00      1.00      1.00      1700\n",
            "weighted avg       1.00      1.00      1.00      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1322    0    0]\n",
            " [   0  231    0]\n",
            " [   1    0  146]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9994117647058823\n",
            "______________________________test______________________________________\n",
            "                     ___RandomForestClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.92      0.96      0.94       333\n",
            "     class 2       0.74      0.58      0.65        64\n",
            "     class 3       0.74      0.79      0.77        29\n",
            "\n",
            "    accuracy                           0.89       426\n",
            "   macro avg       0.80      0.78      0.79       426\n",
            "weighted avg       0.88      0.89      0.89       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[319  10   4]\n",
            " [ 23  37   4]\n",
            " [  3   3  23]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8896713615023474\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_test, Y_predec_model_4_withacp, 'RdBu')  \n",
        "                        ###########################################\n",
        "                        ############___RandomForestClassifier___###\n",
        "                        ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                     ___RandomForestClassifier___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_4_withacp_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_4_withacp_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_RandomForest_withacp = accuracy_score(Y_train, Y_predec_model_4_withacp_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_RandomForest_withacp)\n",
        "print(\"______________________________test______________________________________\")             \n",
        "print(\"                     ___RandomForestClassifier___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_4_withacp, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_4_withacp))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_RandomForest_withacp = accuracy_score(Y_test, Y_predec_model_4_withacp)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_RandomForest_withacp)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "f71c27c1",
      "metadata": {
        "id": "f71c27c1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "df3f560b-5374-421e-cbc9-9d140b0be8ee"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                      ___Support Vector Machine (SVM)___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.87      0.97      0.91      1322\n",
            "     class 2       0.58      0.29      0.39       231\n",
            "     class 3       0.63      0.46      0.53       147\n",
            "\n",
            "    accuracy                           0.83      1700\n",
            "   macro avg       0.69      0.57      0.61      1700\n",
            "weighted avg       0.81      0.83      0.81      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1279   13   30]\n",
            " [ 153   68   10]\n",
            " [  43   37   67]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.831764705882353\n",
            "______________________________Test______________________________________\n",
            "               ___Support Vector Machine (SVM)___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.88      0.97      0.92       333\n",
            "     class 2       0.71      0.39      0.51        64\n",
            "     class 3       0.58      0.48      0.53        29\n",
            "\n",
            "    accuracy                           0.85       426\n",
            "   macro avg       0.73      0.61      0.65       426\n",
            "weighted avg       0.83      0.85      0.83       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[323   4   6]\n",
            " [ 35  25   4]\n",
            " [  9   6  14]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8497652582159625\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_test, Y_predec_model_5_withacp, 'RdBu')  \n",
        "                        ###########################################\n",
        "                        ####___Support Vector Machine (SVM) ___####\n",
        "                        ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                      ___Support Vector Machine (SVM)___\")            \n",
        "print(classification_report(Y_train, Y_predec_model_5_withacp_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_5_withacp_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_SVM_withacp = accuracy_score(Y_train, Y_predec_model_5_withacp_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_SVM_withacp)\n",
        "print(\"______________________________Test______________________________________\")               \n",
        "print(\"               ___Support Vector Machine (SVM)___\")            \n",
        "print(classification_report(Y_test, Y_predec_model_5_withacp, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_5_withacp))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_SVM_withacp = accuracy_score(Y_test, Y_predec_model_5_withacp)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_SVM_withacp)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "dec9a0c7",
      "metadata": {
        "id": "dec9a0c7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "b64c16c3-a2de-4bbe-9c60-e9ec731ef9c3"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                      ______Naives Bayes______\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.91      0.95      0.93      1322\n",
            "     class 2       0.59      0.55      0.57       231\n",
            "     class 3       0.79      0.56      0.65       147\n",
            "\n",
            "    accuracy                           0.86      1700\n",
            "   macro avg       0.76      0.68      0.72      1700\n",
            "weighted avg       0.86      0.86      0.86      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1258   45   19]\n",
            " [ 102  126    3]\n",
            " [  24   41   82]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8623529411764705\n",
            "______________________________Test______________________________________\n",
            "               ______Naives Bayes______\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.93      0.96      0.95       333\n",
            "     class 2       0.67      0.67      0.67        64\n",
            "     class 3       0.85      0.59      0.69        29\n",
            "\n",
            "    accuracy                           0.89       426\n",
            "   macro avg       0.82      0.74      0.77       426\n",
            "weighted avg       0.89      0.89      0.89       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[319  13   1]\n",
            " [ 19  43   2]\n",
            " [  4   8  17]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8896713615023474\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_test, Y_predec_model_6_withacp, 'RdBu')  \n",
        "                        ###########################################\n",
        "                        ##########_______naive bayes _______#######\n",
        "                        ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")               \n",
        "print(\"                      ______Naives Bayes______\")            \n",
        "print(classification_report(Y_train, Y_predec_model_6_withacp_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_train, Y_predec_model_6_withacp_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_naive_Bayes_withacp = accuracy_score(Y_train, Y_predec_model_6_withacp_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_naive_Bayes_withacp)\n",
        "print(\"______________________________Test______________________________________\")               \n",
        "print(\"               ______Naives Bayes______\")            \n",
        "print(classification_report(Y_test, Y_predec_model_6_withacp, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_test, Y_predec_model_6_withacp))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_naive_Bayes_withacp = accuracy_score(Y_test, Y_predec_model_6_withacp)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_naive_Bayes_withacp)"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "d90c2bad",
      "metadata": {
        "id": "d90c2bad"
      },
      "source": [
        "###### 1.3 Without feature selection:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "06c08aea",
      "metadata": {
        "id": "06c08aea",
        "outputId": "c231dca5-4cce-4e2e-9aac-69d2e5ccac97",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___KNeighborsClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       1.00      1.00      1.00      1322\n",
            "     class 2       1.00      1.00      1.00       231\n",
            "     class 3       1.00      1.00      1.00       147\n",
            "\n",
            "    accuracy                           1.00      1700\n",
            "   macro avg       1.00      1.00      1.00      1700\n",
            "weighted avg       1.00      1.00      1.00      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1322    0    0]\n",
            " [   0  231    0]\n",
            " [   0    0  147]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 1.0\n",
            "______________________________Test______________________________________\n",
            "                     ___KNeighborsClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.95      0.96      0.96       333\n",
            "     class 2       0.75      0.73      0.74        64\n",
            "     class 3       0.85      0.79      0.82        29\n",
            "\n",
            "    accuracy                           0.92       426\n",
            "   macro avg       0.85      0.83      0.84       426\n",
            "weighted avg       0.91      0.92      0.91       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[320  12   1]\n",
            " [ 14  47   3]\n",
            " [  2   4  23]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9154929577464789\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_all_test, Y_predec_model_1_without, 'coolwarm')  \n",
        "                        ###########################################\n",
        "                        ############___KNeighborsClassifier___#####\n",
        "                        ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")              \n",
        "print(\"                     ___KNeighborsClassifier___\")            \n",
        "print(classification_report(Y_all_train, Y_predec_model_1_without_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_train, Y_predec_model_1_without_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_KNN_without = accuracy_score(Y_all_train, Y_predec_model_1_without_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_KNN_without)\n",
        "print(\"______________________________Test______________________________________\")              \n",
        "print(\"                     ___KNeighborsClassifier___\")            \n",
        "print(classification_report(Y_all_test, Y_predec_model_1_without, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_test, Y_predec_model_1_without))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_KNN_without = accuracy_score(Y_all_test, Y_predec_model_1_without)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_KNN_without)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "0681a2c6",
      "metadata": {
        "id": "0681a2c6",
        "outputId": "b1b0d419-e936-4780-ae05-a2c3e162051b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___DecisionTreeClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.99      1.00      0.99      1322\n",
            "     class 2       0.98      0.94      0.96       231\n",
            "     class 3       1.00      0.99      1.00       147\n",
            "\n",
            "    accuracy                           0.99      1700\n",
            "   macro avg       0.99      0.98      0.98      1700\n",
            "weighted avg       0.99      0.99      0.99      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1319    3    0]\n",
            " [  15  216    0]\n",
            " [   0    1  146]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9888235294117647\n",
            "______________________________Test______________________________________\n",
            "                     ___DecisionTreeClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.96      0.95      0.95       333\n",
            "     class 2       0.77      0.83      0.80        64\n",
            "     class 3       0.93      0.97      0.95        29\n",
            "\n",
            "    accuracy                           0.93       426\n",
            "   macro avg       0.89      0.91      0.90       426\n",
            "weighted avg       0.93      0.93      0.93       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[315  16   2]\n",
            " [ 11  53   0]\n",
            " [  1   0  28]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9295774647887324\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_all_test, Y_predec_model_2_without, 'coolwarm')  \n",
        "                        ###########################################\n",
        "                        ############___DecisionTreeClassifier___###\n",
        "                        ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")              \n",
        "print(\"                     ___DecisionTreeClassifier___\")            \n",
        "print(classification_report(Y_all_train, Y_predec_model_2_without_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_train, Y_predec_model_2_without_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_DecisionTree_without = accuracy_score(Y_all_train, Y_predec_model_2_without_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_DecisionTree_without)\n",
        "print(\"______________________________Test______________________________________\")               \n",
        "print(\"                     ___DecisionTreeClassifier___\")            \n",
        "print(classification_report(Y_all_test, Y_predec_model_2_without, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_test, Y_predec_model_2_without))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_DecisionTree_without = accuracy_score(Y_all_test, Y_predec_model_2_without)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_DecisionTree_without)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "109e6806",
      "metadata": {
        "id": "109e6806",
        "outputId": "e34eafde-a218-4c05-fe59-aefe0a7a7403",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___LogisticRegression___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.95      0.96      0.95      1322\n",
            "     class 2       0.69      0.68      0.69       231\n",
            "     class 3       0.84      0.82      0.83       147\n",
            "\n",
            "    accuracy                           0.91      1700\n",
            "   macro avg       0.83      0.82      0.82      1700\n",
            "weighted avg       0.91      0.91      0.91      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1263   52    7]\n",
            " [  57  158   16]\n",
            " [   7   20  120]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9064705882352941\n",
            "______________________________Test______________________________________\n",
            "                     ___LogisticRegression___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.94      0.93      0.94       333\n",
            "     class 2       0.60      0.58      0.59        64\n",
            "     class 3       0.65      0.76      0.70        29\n",
            "\n",
            "    accuracy                           0.87       426\n",
            "   macro avg       0.73      0.76      0.74       426\n",
            "weighted avg       0.87      0.87      0.87       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[310  20   3]\n",
            " [ 18  37   9]\n",
            " [  2   5  22]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8661971830985915\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_all_test, Y_predec_model_3_without, 'coolwarm')  \n",
        "                        ###########################################\n",
        "                        ############___LogisticRegression___#######\n",
        "                        ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")              \n",
        "print(\"                     ___LogisticRegression___\")            \n",
        "print(classification_report(Y_all_train, Y_predec_model_3_without_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_train, Y_predec_model_3_without_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_LogisticRegression_without = accuracy_score(Y_all_train, Y_predec_model_3_without_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_LogisticRegression_without)\n",
        "print(\"______________________________Test______________________________________\")             \n",
        "print(\"                     ___LogisticRegression___\")            \n",
        "print(classification_report(Y_all_test, Y_predec_model_3_without, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_test, Y_predec_model_3_without))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_LogisticRegression_without = accuracy_score(Y_all_test, Y_predec_model_3_without)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_LogisticRegression_without)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "c3aba592",
      "metadata": {
        "id": "c3aba592",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "b29a6efe-8747-45b1-d94c-0960a8626d14"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___RandomForestClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       1.00      1.00      1.00      1322\n",
            "     class 2       1.00      1.00      1.00       231\n",
            "     class 3       1.00      1.00      1.00       147\n",
            "\n",
            "    accuracy                           1.00      1700\n",
            "   macro avg       1.00      1.00      1.00      1700\n",
            "weighted avg       1.00      1.00      1.00      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1322    0    0]\n",
            " [   0  231    0]\n",
            " [   0    0  147]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 1.0\n",
            "______________________________Test______________________________________\n",
            "                     ___RandomForestClassifier___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.96      0.98      0.97       333\n",
            "     class 2       0.88      0.78      0.83        64\n",
            "     class 3       0.90      0.90      0.90        29\n",
            "\n",
            "    accuracy                           0.94       426\n",
            "   macro avg       0.91      0.89      0.90       426\n",
            "weighted avg       0.94      0.94      0.94       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[326   6   1]\n",
            " [ 12  50   2]\n",
            " [  2   1  26]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9436619718309859\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_all_test, Y_predec_model_4_without, 'coolwarm')  \n",
        "                        ###########################################\n",
        "                       ############___RandomForestClassifier___###\n",
        "                       ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")              \n",
        "print(\"                     ___RandomForestClassifier___\")            \n",
        "print(classification_report(Y_all_train, Y_predec_model_4_without_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_train, Y_predec_model_4_without_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_RandomForest_without = accuracy_score(Y_all_train, Y_predec_model_4_without_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_RandomForest_without)\n",
        "print(\"______________________________Test______________________________________\")               \n",
        "print(\"                     ___RandomForestClassifier___\")            \n",
        "print(classification_report(Y_all_test, Y_predec_model_4_without, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_test, Y_predec_model_4_without))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_RandomForest_without = accuracy_score(Y_all_test, Y_predec_model_4_without)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_RandomForest_without)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "92aeaae8",
      "metadata": {
        "id": "92aeaae8",
        "outputId": "157c2d9e-cdd1-49a7-cd51-7e9cbcf09694",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAqsAAAJ8CAYAAADZIOhmAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABl4ElEQVR4nO3deVhV5drH8d/eDIIiIIoozpJjoKE54TykFTlkVmrmEMfZTMzCcqjzpqWpZTnk0KRHT5lmpmWDHsshzMop1NLUTMQRBRFEpr3fPzhxIhRBca+15fu5rn0pay3Wc6/VDm/ufT/PstjtdrsAAAAAE7IaHQAAAABwLSSrAAAAMC2SVQAAAJgWySoAAABMi2QVAAAApkWyCgAAANMiWQUAAIBpkawCAADAtFyNDgAAAKA4OnjwoNLT0w0Z293dXXXq1DFk7MIiWQUAADBAenq60i6nKuvMeYeO6xJQ1qHj3SySVQAAAINknTmv+IfGOHTMch/PlmpUduiYN4NkFQAAwCgWyeJmcfiYzoQJVgAAADAtklUAAACYFm0AAAAABrFIsro69nN5J+sCoLIKAAAA86KyCgAAYBSLZHFzcO3QyUqrVFYBAABgWiSrAAAAMC3aAAAAAAzk6AlWzobKKgAAAEyLyioAAIBReILVdVFZBQAAgGlRWQUAADCIxWLAQwGorAIAAABFg2QVAAAApkUbAAAAgGEsjp9g5WQzrKisAgAAwLRIVgEAAIzy3wlWjnzdSGH1119/1YABA9S4cWOFhYVpzJgxOnfunCRp+/bt6tWrlxo1aqTw8HCtXbs21/cuXbpUXbp0UaNGjdSnTx/t27evUGOTrAIAAOCa0tPT9cQTT6hp06bavn27PvvsM50/f14vvviizp49qxEjRqh3797avn27JkyYoEmTJikmJkaStGnTJs2ZM0evvvqqoqOj1b59ew0bNkyXL18u8PgkqwAAALim1NRURUZGaujQoXJ3d5efn5/uuece/fbbb1q3bp2qV6+uXr16qUSJEgoLC1OHDh20cuVKSdKKFSvUs2dPNWzYUB4eHvrHP/4hSfrmm28KPD7JKgAAgIEsLhaHvgrLx8dHDz/8sFxds+flHz16VJ988onuu+8+7d+/X/Xr1891fP369XM+6v/7fqvVqnr16uVUXguCZBUAAADXFRcXp+DgYN1///0KCQnR6NGjlZiYKG9v71zH+fr6KiEhQZKUmJgoHx+fXPt9fHxy9hcES1cBAAAYxSJZb6DaebNj3ohKlSopJiZGf/zxhyZPnqxnn322QN9nt9tvbMD/orIKAACAArFYLKpevboiIyP12WefydXVVYmJibmOSUhIkJ+fnySpTJkyefYnJibm7C8IklUAAAADWawWh74Ka/v27erSpYtsNlvONqs1O4Vs0KBBnqWo9u3bp4YNG0qSgoODtX///px9WVlZOnDgQM7+giBZBQAAwDUFBwcrOTlZM2bMUGpqqi5cuKA5c+bo7rvvVp8+fRQXF6eVK1cqLS1Nmzdv1ubNm/XII49Ikvr06aM1a9Zoz549Sk1N1VtvvSV3d3e1a9euwOOTrAIAAOCaSpcurXfffVf79u1T8+bNFR4ertKlS+u1115T2bJltXDhQi1btkyNGzfWyy+/rBkzZqhu3bqSpDZt2mjs2LEaM2aMmjZtqujoaC1atEgeHh4FHt9iv9muVwAAABRaTEyM0uLilD5xkkPHdZ/ykkpUqqSQkBCHjnujqKwCAADAtFi6CgAAwCgWiwFLVzl4vJtEZRUAAACmRbIKAAAA06INAAAAwEA3svZpcUJlFQAAAKZFZRUAAMBADp9g5WSorAIAAMC0SFYBAABgWrQBAAAAGMUiWRy+zqpjh7tZVFaBYmr8+PGqU6eOxo8ff81jRowYoTp16mjOnDm3JIbVq1erTp06OnLkyC05//U88sgjevzxxw0Z+0bt3LlTXbp0UXBwsD777LMiO2+dOnU0c+bMIjsfABQVKqtAMVayZEl99dVXmjRpkkqVKpVr34ULF7RlyxZ5enoW+ryxsbHq1KmTDh48mO9x999/v1q3bi0/P79Cj1FcLViwQCkpKfr0009VoUKFIjvvtm3bVLJkySI7H4CCsUiSldphfrg7QDFWr149ubq66osvvsiz77PPPlO1atVuKJHcvXt3gY7z8PCQv7+/XFxcCj1GcZWUlKQaNWooKCgozy8YN8Pf379IzwcARYVkFSjGXFxc1LZtW61evTrPvjVr1qhDhw55tp87d07jx49XixYtFBwcrA4dOmjatGm6cuWKJGnOnDl65plnJClXm0GdOnW0aNEiDR06VCEhITp48OBV2wA2b96sXr16qUGDBmrXrp1eeuklJScn5xr/2WefVYcOHRQSEqLw8HCtWrXqute6ceNG3XvvvQoODta9996r9evX5zkmPT1db7zxhsLDw9WgQQO1bdtWM2fOVHp6er7nvnTpkl588UW1bNlSoaGhevTRR/Xdd9/lOmb16tXq2rWrQkJC1LhxY0VERGjfvn259tepU0eHDh3S4MGDFRoaqlatWunll1+WzWbLuYd79uzRDz/8oDp16mj16tWaM2eO6tSpo7S0tFzj/fVjfbvdrgULFqhLly5q0KCBmjdvrlGjRik2Nvaqx0vSmTNn9PTTT6t58+YKDg5Wp06d9OabbyozMzPnmA4dOmjq1Klavny5OnbsqLvuuku9evXSzz//fL3/HAD+wmK1OPTlbEhWgWKua9eu2rlzp44dO5az7dChQ9q/f7/Cw8PzHP/000/rp59+0vz587Vhwwa98MIL+vjjjzV79mxJ0hNPPKF+/fpJyv5oecKECTnfu3LlSjVu3FhffPGFatSokefcP/30k4YNG6aWLVvqk08+0SuvvKKvv/465xzp6ekaMGCAdu7cqRdffFHr1q1T9+7dNXHiRK1Zs+aa13j06FE99dRTCgoK0urVqzV9+nStWLFCx48fz3XcP//5T73zzjsaMGCAPvvsM0VFRWnlypV64YUX8r2HY8aM0XfffaeZM2dqzZo1CgkJ0dChQ3XgwAFJ0qpVq/Tcc8+pU6dOWrNmjd5//31lZGSof//+On36dK5zvfjii3r44Ye1du1aPfroo1qyZElO5Xvbtm268847FRoaqm3btun+++/PN64/rVq1SgsXLtQzzzyjL7/8UosWLVJSUpKGDh161ePT0tLUv39//fLLL3rttde0fv16/eMf/9DixYs1Y8aMXMdu3bpVe/fu1YIFC7R06VJdvHhRzz77bIHiAoCCoGcVKOZatmypsmXLavXq1Ro7dqwk6ZNPPlHt2rVVt27dPMdPmzZNFotFFStWlCRVrFhRrVq10tatWzV+/HiVKlUqp8/V398/1/eWLl1aQ4YMuWYsb7/9tmrXrq3IyEhJUlBQkCZOnKjNmzcrIyNDGzdu1JEjR7RkyRI1b95ckjRkyBDt2bNHb731lnr06HHV83766aeyWCyaNm2aSpcuLUl67bXX1Lp165xjzpw5o9WrV2v48OF65JFHJElVq1bV2bNnNW3aNI0ZM0YBAQF5zr1v3z5t27ZN8+bNU4sWLSRJzz33nJKSknTy5EnVr19fixcvVps2bfTUU0/lfN9rr72mNm3aaPXq1RoxYkTO9vvvv1+dO3eWJA0fPlwLFy7Uzz//rPDwcPn7+8vV1VVubm557m1+9u/fr4oVK6pTp06SpMDAQM2ePVsnT56UzWaT9W/9chs2bNCxY8e0cuVKNWjQIOde/Pbbb1qxYoWefvppubu7S5KSk5M1ZcqUnK+7d++uOXPmKDk5WV5eXgWOEQCuhcoqUMy5urrq/vvv15o1a2Sz2ZSVlaV169apa9euVz0+IyNDc+fO1T333KPGjRsrNDRUX3/9tRITE687VnBwcL77f/75ZzVs2DDXti5duujll1+Wm5ub9u7dKzc3NzVt2jTXMS1atNCxY8eUkpJy1fP+9ttvqlq1ak6iKklly5ZV1apVc77et2+fbDabWrZsmefcdrs9p0p6tZgl5SR1UnZ7xauvvqpOnTopOTlZx44d0913353r+8qVK6cqVarkOe9fr9/V1VXe3t5KSkq66tgF1b59ex07dkwDBw7UJ598olOnTsnPz0/BwcF5ElVJiomJUYkSJRQSEpJre2hoqFJTU3X06NGcbXfeeWdOoiopp8f54sWLNxUzUGxYsp9g5ciXsy1dRWUVgLp166Z//etf2rZtm+x2u+Lj4/XAAw/kOS4lJUX9+vWTm5ubnnnmGdWqVUtubm6aOXOmdu3add1xvL29892flJSU7ySf5ORkZWRkqHHjxrm2/9lHee7cuat+f0pKylVnuv/12D/7Yp944olcCZzdbs8599VcunQpz7n+HrOkq1YZvby88iTYf4/TYrHkxHCj2rZtq6VLl2rp0qWaOnWqLl26pIYNGyoqKirPvfwz5lKlSsliyf0v2p/X8NeYrxavpJuOGQD+RLIKQA0aNFCNGjW0fv36nGQwMDAwz3E7duzQ2bNn9fbbb+f6CP3y5ctFEkfZsmXzrch5e3vLw8Pjmv2pf7Ym/J2np6fOnz+fZ/ulS5dyki0fHx9J0syZM1W7du08x15rVYQ/t18r0f4zwfvrJLE/JScnq1KlSlc9b0FdLTm8WoX57rvv1t13363MzEzt3LlTc+fO1eDBg/Xtt9/m+SXC29tbKSkpstvtuRLWPxPz6/3SAaAwjJj05FylVdoAAEjKrq5u27ZNW7ZsybcFQMqduJ04cUI7duy4aiWtsNW12rVr66effsq1bcOGDXrssceUkpKiu+66S1euXFFqaqqqVauW8/Lw8JC3t3euj6P/KigoSMeOHcuVCJ85cybXbPjg4GC5uLjo5MmTuc7t7+8vq9Waq4Xgr+rUqSNJ+uGHH3JtHzZsmP71r3/Jy8tLd9xxh3788cdc+8+ePavY2Ng8H7UX1p9xXbhwIWfb3r17cx2zdetWHT58WFJ2a0GzZs303HPPKSUlJdc9+FODBg2UlpaWZ1b/zp075eXlperVq99UzABQGCSrACRlJ6vx8fFKTU3Vvffee9VjgoOD5erqqnfffVexsbHavn27Ro4cqfvuu0+JiYk6cOCA0tPTcypvGzduzNXfeD0RERGKjY3VSy+9pNjYWH3//fd65ZVX5Ofnp1KlSql9+/aqXbu2xo0bp+joaMXFxWnz5s3q16+fJk2adM3zdu3aVVlZWXrxxRd1+PBh7d27V+PGjVPZsmVzjilXrpx69eqluXPnas2aNYqNjdXevXs1evRo9evXT6mpqVc9d4MGDdSsWTPNmDFDO3bs0PHjxzV9+nRt27ZNjRo1kiQNHjxYW7du1dy5c3Xs2DHt2bNHTz31lHx9ffXQQw8V+P5ca3wp+2EBx48f1/bt2zVnzpxcbQerV6/WyJEjtW3bNp08eVKHDh3Se++9p7JlyyooKCjPOTt27KigoCA9//zz+uGHH3T8+HH961//0qpVqzRo0CC5ubndVMwAUBi0AQCQJFWuXFmNGzeWt7e3fH19r3pMpUqVNHXqVL355pt64IEHVLt2bU2ePFllypTRjz/+qMcee0wrV65Ut27dtG7dOo0ZM0bt27fX3LlzCxRD8+bNNW/ePM2dO1cfffSR/Pz81KlTp5zVAdzd3fX+++9r5syZevrpp3Xx4kWVK1dO4eHhGj169DXPW7duXc2YMUNvvPGGevTooUqVKumpp57Sxx9/nGsN1cmTJ6t8+fKaM2eOTp8+rVKlSqlVq1ZatmxZvk/ymjt3rmbMmKExY8YoNTVVtWrV0sKFC3XnnXdKknr06CGbzab33ntPCxYskIeHh5o2baqpU6fe9NO7QkNDFRkZqeXLl2vNmjWqV6+eJk2alGtZqpdeekkzZ87UhAkTdP78eXl7e6thw4Z699135eHhkeec7u7ueu+99zR9+nQ9+eSTSklJUaVKlTRu3DgNGDDgpuIFkJeFJ1jly2KnCx4AAMDhYmJilH76lNxef9mh42ZEPi/3ChVvug3JUaisAgAAGMgZnyrlSNSdAQAAYFpUVgEAAAxi+e9DARw9pjOhsgoAAADTIlkFAACAadEGAAAAYCAmWOWPZLWI7N69W3a7ncWyAQBwIhkZGbJYLAoNDTU6FFwDyWoRsdvtsmVk6vKJM0aH4tQsLlZZA8rKdua87Fk2o8Nxekle5Y0O4bZgtUo+XtLFZMnG2/Km2biJN83FKvl6W5SYZBc/Km9OGW+L4f9f81CA/JGsFhE3NzddPnFG53qOMToUp+Zap7r8l76sC8++psyDx4wOx+nNe3yl0SHcFir62zXy0UwtX++qU+f4uO5mJcUnGB2C06tc0apnh5bU4g9TdeIU2erNmPxUSaNDwHWQygMAAMC0qKwCAAAYxWJx/AQrJ1tolcoqAAAATIvKKgAAgIFYuip/VFYBAABgWlRWAQAADERlNX9UVgEAAGBaJKsAAAAwLdoAAAAADMQTrPLH3QEAAIBpUVkFAAAwiMUiWV0cO8HKyZ4JQGUVAAAA5kWyCgAAANOiDQAAAMBArLOaPyqrAAAAMC0qqwAAAIaxGLB0lXNVcqmsAgAAwLSorAIAABjFYkDPqnMVVqmsAgAAwLxIVgEAAGBatAEAAAAYiKWr8kdlFQAAAKZFZRUAAMBAjl+6yrlwdwAAAGBaJKsAAAAwLdoAAAAADMQEq/xRWQUAAIBpUVkFAAAwiMVicfgEK4vFuSq5VFYBAABgWlRWAQAAjORklU5Ho7IKAAAA0yJZBQAAgGnRBgAAAGAglq7KH5VVAAAAmBaVVQAAAAM5eukqZ8PdAQAAgGmRrAIAAMC0aAMAAAAwisWACVZONp+LyioAAABMi8oqAACAYSwGTLByrtIqlVUAAACYFpVVAAAAA/FQgPxRWQUAAIBpkawCAADAtGgDAAAAMBBtAPmjsgoAAADTorIKAABgFIskRy9d5WSFXCqrAAAAMC2SVQAAAJgWbQAAAAAGslic7HN5B6OyCgAAANOisgoAAGAQiySLgydYOVsdl8oqAAAATItkFQAAAKZFGwAAAIBhLAY8wcq5GgGorAIAAMC0qKwCAAAYhSdYXReVVQAAAOQrLi5OI0eOVLNmzRQWFqbx48crKSlJJ06cUJ06dRQSEpLr9c477+R87/r169W1a1eFhoaqZ8+e2rZtW6HGprIKAABgIMf3rBbesGHDFBwcrE2bNunSpUsaOXKkpk+fruHDh0uSYmJirvp9v/zyi6KiojR37lw1b95cX331lUaNGqUvv/xSFSpUKNDYVFYBAABwTUlJSQoODtbTTz+tUqVKqUKFCnrwwQf1008/Xfd7V65cqbZt26pt27YqUaKEunXrptq1a2vt2rUFHp9kFQAAANfk7e2tV155ReXKlcvZdurUKZUvXz7n62effVatWrVS8+bNNWvWLGVkZEiS9u/fr/r16+c6X/369a9Zib0aklUAAAADWSxWh75uVkxMjJYtW6bhw4fL3d1doaGhuueee/TNN99o0aJFWrt2rebPny9JSkxMlI+PT67v9/HxUUJCQoHHI1kFAABAgezcuVMRERF6+umnFRYWpvLly+vDDz/UPffcIzc3NzVo0EBDhw7V6tWrc77Hbrff1JhMsAIAADCSE0ywkqRNmzbpmWee0aRJk9SjR49rHlepUiXFx8fLbrerTJkySkxMzLU/MTFRfn5+BR6XyioAAADytWvXLkVFRemNN97Ilahu375db731Vq5jjx49qkqVKslisSg4OFj79u3LtT8mJkYNGzYs8NgkqwAAALimzMxMTZw4UePGjVOrVq1y7StdurTmzZunTz/9VBkZGYqJidE777yjPn36SJIeeeQRRUdH69tvv1VaWppWrVqlY8eOqVu3bgUenzYAAAAAw1hkcfQTrAr5CKs9e/boyJEjmjJliqZMmZJr35dffqnXX39dc+fO1eTJk1W6dGk9/vjjGjBggCSpdu3amjlzpl555RXFxcXpjjvu0MKFC+Xv71/g8UlWAQAAcE133323Dh48eM39lSpV0j333HPN/Z07d1bnzp1veHySVQAAAKNYDHiClXPM58pBzyoAAABMi8oqAACAkYpgof7bGXcHAAAApkWyCgAAANOiDQC3nsWimmMGqvKAnioVVFVZKamK/+Z7/RL1qlKPn8w5rFzHFgpd/pqsPqX1n6ucxrNqoOpOHSu/1k3k6u2llINHdfjVxTr9ydeOuxbc9np0LqcnHq6g6J1JmrbgeK59Det5a9yQ8qoa6CGrRfrlyGW9v+q0Dv+RalC0KG4sFunR7pXV494KqlTRU4tfy9KPexI0792jOnMuzejwcIMcPsHKyVBZxS1Xf0aUak9+UkdeXaxvQ8K1q99Y+TS6U803LpXFzU2yWlX7n0+pySdvKSv16j9s3Xy91eKbZSp1RzXt6hupbc17KfGHn3X3yjkqd09LB18RbkdepVz0wujqeuhef6Wn532Otbe3t54cUF0//5qiMf/3m8a9ckRpaTZNj6qpgHJuBkSM4mjUE0GK6FNNa76M1f79+/Xm4l9VN6i05rzcUK6uJDy4PZkyWT1x4oTq1KmjI0eOGB0KbpLFxUUVHuysIzPfVty/1yr12AnFb4zWoX/OUamaVeTdoI5K1wtS5ce6Kbp9PyXsjLnqeQIfDVfJapW0s/cYJUTvUsqh37VvzBSlxp5S9aF9HHxVuB21b+4rDw+rRr1wSMmXs/LsL1u2rM4npGvJx6cVdyZdv8de0ZylcSrp6aIWjXwMiBjFjYtVahdWTstXx2rbjnNKT09XzC+Jevvfx1SpgqfuqF7K6BBxo6xWx76cDG0Af5GSkqIXXnhB69at0/r16xUUFGR0SE7PnpWlTUEd8m632SRJtowMXYk7o61Neioj4eI1z3P83VU6s+4/unLy7F9OYlfa6XNy8SpZ5HGj+Plhb5I+33RetrxF1Rx/35eRYbu1QQF/kWWTHorYIUmqXPF/CYf9v+/LzMx83ryAE3O+9PoWOXPmjHr27CkXFxejQ7nted9VT7UmjNDpdZt06eeDykhMyjdRlSR7RkbuRFWSR5WK8m5YVwk79t7KcFFMnInPyDdRPXfunMqVcVe3TmVltUhurhY9/mAFXUrO1JYfEh0WJ/BX1auU0sBHq2rbjngdPpZidDjALWFoshobG6snnnhCoaGhat++vZYuXXrV444fP66IiAg1a9ZMzZo109ixY5WUlCRJstlsmjZtmlq1aqW77rpL3bp109atWyVJqampioqKUosWLRQaGqrevXtr3759Vx0jISFBzzzzjJ588slbc7FQ3VfG6f7LMWq942PFb/xOOx++8XttLeGuRstmKf18on5/Y0kRRglcXXJyshb8+w8NfKiCPl0Uok8WBqtlY289P/N3XUjMNDo8FDN9e1ZXaGioXpkQqh/3JOj5Vw4YHRJukMViMeTlTAxNVkeNGqWgoCBFR0dr/vz5mj17tr777rs8x02cOFHly5fX1q1b9cUXX+j333/X/PnzJUmff/65oqOjtXbtWu3cuVMDBgxQVFSUMjIytGTJEsXHx2vDhg3asWOHWrdurUmTJl01lrp166pTp0639HqLuyMz39GWxj20Z2CUKnTvpCZrF9xQ74xr6VJqtv5tlb6zln7sMVwZFxKLPljgb7y8vBTxSBV9ufmCnn75sJ6fcVSHfk/VC6OrMcEKDrfu6xP65ZdfNO/dg2rdvJxmTA52xlZEoEAM61k9cOCADh48qCVLlsjT01P16tXT3LlzFRAQkOfYRYsWyWKxyN3dXX5+fmrdurV27dolSUpKSpKrq6s8PT3l4uKihx56SA8++KCsVquSkpLk5uYmDw8Pubq6asSIERoxYsQtuyaLi1WudarfsvM7O7ukK7YsXdkVo9RxLyvso7mqPKqfTn+1JecYS6ns/lPXaoFXPUeJcn5qvHCK3Mv46IcnopSSnMw9z0dFf3rYboSLVfIoYc+5f+XK2FW5cmXFnrysdf/JXm7tUrL03qoUzXy+ngY85K9/rY4zMmSn4u1GVnWzSnpk6cqVK/rtaKreWJSqVyaEquf95fX9znijQ3M6ri5SZt45lY7Fbxr5MixZPX78uLy8vOTr65uzLSwsTFL2agB/tW/fPs2aNUsHDx5URkaGsrKyFBwcLEkKDw/Xp59+qjZt2qhly5Zq166dwsPDZbVa1bdvX0VERKht27Zq3bq1OnXqpI4dO96ya7IGlJX/0pdv2fmdkYuLi7y9vXXp0iVlZv7vo9I/P4IoP26Qsh67N2e7e7VqkqQyL4266rnq1KkjSTr022/ymPAPedzK4G8DI8XH0zeidCm7alW1a+Sj/7t/Hh4e8i4Zn2ubJLlY0tS4vru83bjXBcekyBtxtZ+nAx7ylMWS/UvVo1191Obuy0aG6LTiE5gsaWaGJatWq1U22/XfHBcvXtSQIUPUp08fLV68WF5eXpo9e7aio6MlSb6+vvroo4+0a9cuffPNN3rzzTf1wQcfaPny5apcubLWr1+vHTt2aNOmTZo8ebLWrl2rN99885Zck+3MeV149rVbcm5n5VHBX3f9Z5kOznpbv7+7Mme7T0gd6cNGurDwI51b9UXO9oqzJ0ntmilh0lxl/nEy17mavDtdtrMXtKNfpNIv5D8hC9k+Cn/V6BCc0qs1LDpy3KKFK7J/RJYrY1f9+um6eNlT81b878emi4tFr03w0N5fL+q9VSyuUlApiUlGh+CUypZx1/zpNbX849+1Y2ecBjzkqSUfp6p0aS81aiSt25iiTdtIVgtrSB/jyx48FCB/hv10rVKlilJSUnT27FmVL19ekrRx40Z5e3srMPB/HwEfPXpUKSkpioiIkJeXl6TsFoI/paWlyWazqVGjRmrUqJGGDx+uli1b6tdff1WNGjXk5uamsLAwhYWFadCgQerQoYMSEhJUpkyZIr8me5ZNmQePFfl5nVnywWOKff9j1RzcW6kHjujC1h/lWa2S7pwwUldOndWJuctkz8yS63+Xn7L897MYa0qqXBKTlZV6RZlJyQp85H6VbXaXfnxwhLKOxunvazakneGjr6s51ZQfgAXlVcpFbi7Z98suKctm1ZW07F7Ui5cydfbsWdWuUU3twwK0ZcdFublZ1LOLv0p6uujTjQk6dY57XVBJ8VSxbsSJU1f0+cbT6nFfFSVdypC7e5rKlXXTYw/VVPyFNK1ef0bJKdzbwjK8BQDXZViyWq9ePdWvX1+zZ8/WhAkTFBcXpwkTJmjGjBm5jgsMDJTVatXu3bvVokULffTRR4qPj1diYqIyMzM1depUJSYm6sUXX1SZMmW0f/9+2Ww2BQYGavTo0apevboiIyNVsmRJ7d69W76+vvLxYQFvR4oZ8YKunDyrWhNGyLNygNJOx+vCtp36ddLryrx4SbUnj1LtyblXBuiw+UNJUuyS1dob8Zz8O7eSJDX5ZP5Vx/jMtc6tvQjc9iaNqqYGdb1yvvb3c1fYfxf7f2fFccXHn9PmnS5q3bSseoeXV5ZNOvJHqia//rv2/0Y1C44xY94hnTufpp7hVeRftoQqVkrXrp8vauG/fldyClkXbk8Wu91u2AyMM2fO6Nlnn9WePXtUtmxZDRgwQAMGDNCJEyfUsWPHnIX5Fy9erLfffluS1LdvX3Xt2lX9+/dXYGCgFi9erBdeeEHfffedMjMzVa1aNY0aNUqdOnXSyZMnNXnyZO3atUsWi0W1atXSuHHjdPfdd+eJZf78+Xrrrbdkt9uVkZEhNzc3WSwWDR8+vECTsmJiYnT59xM613NMUd+mYsW1TnX5L31Z5/o/T5W6CMx7fOX1D8J1VfTP7l+dt8KVCmoRSIpPMDoEp1e5olXPDi2pVxde1olTVFNvxuSnsj/Z69KpicPHjomJkT3pgmru/eL6Bxehow3vk8XbTyEhIQ4d90YZ2mQVEBCgJUvyrpFZuXJlHTx4MOfrwYMHa/DgwbmO2bZtW87fZ8+efdXzBwYG5iS513OrVwoAAABA4TEjAAAAwEhMsMoXC3sBAADAtEhWAQAAYFq0AQAAABjFYpHF4uDaocW52g6orAIAAMC0qKwCAAAYiQlW+aKyCgAAANOisgoAAGAgi5XaYX64OwAAADAtklUAAACYFm0AAAAARnKypaQcjcoqAAAATIvKKgAAgFEskhw9wcrJCrlUVgEAAGBaJKsAAAAwLdoAAAAADGMxYIKVc/UBUFkFAACAaVFZBQAAMBBPsMofdwcAAACmRWUVAADASBZqh/nh7gAAAMC0SFYBAABgWrQBAAAAGMnqXEtJORqVVQAAAJgWlVUAAAADWZhglS/uDgAAAEyLZBUAAACmRRsAAACAUSwWx0+wsjjXhC4qqwAAADAtKqsAAABGYoJVvrg7AAAAMC0qqwAAAEZysh5SR6OyCgAAANMiWQUAAIBp0QYAAABgJCu1w/xwdwAAAGBaVFYBAAAMYzFg6SrnmtBFZRUAAACmRbIKAAAA06INAAAAwCgWSVYHfyzvXF0AVFYBAABgXlRWAQAAjOTwCVbOhbsDAAAA0yJZBQAAgGnRBgAAAGAki5PNeHIwKqsAAAAwLSqrAAAARrJSO8wPdwcAAACmRWUVAADASPSs5ovKKgAAAEyLZBUAAACmRRsAAACAYSwGPMHKudoOqKwCAADAtKisAgAAGMUixy9d5VyFVSqrAAAAMC+SVQAAAJgWbQAAAABGYp3VfFFZBQAAgGlRWQUAADCSw5euci7cHQAAAJgWlVUAAADDWAzoWXWuHlkqqwAAADAtklUAAACYFm0AAAAARnL0E6ycDHcHAAAApkVlFQAAwCgWye7oCVbONb+KyioAAADMi2QVAAAApkUbAAAAgJF4glW+uDsAAAAwLZJVAAAAI1msjn3dgLi4OI0cOVLNmjVTWFiYxo8fr6SkJEnSL7/8on79+qlx48bq3Lmz3n333Vzfu379enXt2lWhoaHq2bOntm3bVqixSVYBAACQr2HDhsnb21ubNm3S6tWr9dtvv2n69Om6cuWKhg4dqubNm2vr1q16/fXXtXDhQn399deSshPZqKgojRs3Tt9//70GDhyoUaNG6fTp0wUem2QVAADAMBbZLY59FXbtqqSkJAUHB+vpp59WqVKlVKFCBT344IP66aef9O233yojI0PDhw9XyZIldeedd+rhhx/WihUrJEkrV65U27Zt1bZtW5UoUULdunVT7dq1tXbt2gKPT7IKAACAa/L29tYrr7yicuXK5Ww7deqUypcvr/3796tOnTpycXHJ2Ve/fn3t27dPkrR//37Vr18/1/nq16+vmJiYAo9PsgoAAIACi4mJ0bJlyzR8+HAlJibK29s7135fX18lJibKZrMpMTFRPj4+ufb7+PgoISGhwOOxdFURSiodoAWDVhsdhlOrWM6mYUrXx91m6lQ8v0vdLP+KfkaHcFso45MuKV5lAnyV6eFudDhOLym+4P9IAcWCEy1dtXPnTg0fPlxPP/20wsLC9MUXX1z1OMtfnsplt9tvakznuTsAAAAwzKZNmzRkyBA9//zz6t+/vyTJz88vT5U0MTFRvr6+slqtKlOmjBITE/Ps9/MreDGFZBUAAMBIFotjXzdg165dioqK0htvvKEePXrkbA8ODtbBgweVmZmZsy0mJkYNGzbM2f9n/+rV9hcEySoAAACuKTMzUxMnTtS4cePUqlWrXPvatm0rLy8vvfXWW0pNTdXevXu1atUq9enTR5L0yCOPKDo6Wt9++63S0tK0atUqHTt2TN26dSvw+PSsAgAA4Jr27NmjI0eOaMqUKZoyZUqufV9++aUWLFigF154QYsWLVK5cuUUGRmpdu3aSZJq166tmTNn6pVXXlFcXJzuuOMOLVy4UP7+/gUen2QVAADASFZzf9B999136+DBg/ke88EHH1xzX+fOndW5c+cbHt/cdwcAAADFGpVVAAAAo1j036dKOXZMZ0JlFQAAAKZFZRUAAMAwFgMeCuBcpVUqqwAAADAtklUAAACYFm0AAAAABrI7vA3AuXB3AAAAYFpUVgEAAIzk6KWrnAyVVQAAAJgWySoAAABMizYAAAAAAzHBKn/cHQAAAJgWlVUAAAAjMcEqX1RWAQAAYFpUVgEAAAxjkRzes+pclVwqqwAAADAtklUAAACYFm0AAAAABrFbJLuDP5a3y7kaAaisAgAAwLSorAIAABjJ0ROs7I4d7mZRWQUAAIBpkawCAADAtGgDAAAAMJCjJ1g5GyqrAAAAMC0qqwAAAIaxyO7wCVbOVcmlsgoAAADTorIKAABgJEdXVp0MdwcAAACmRbIKAAAA06INAAAAwEB2i3NNeHI0KqsAAAAwLSqrAAAABnL40lVOhrsDAAAA0yJZBQAAgGnRBgAAAGAkJljli8oqAAAATIvKKgAAgGEsBkywcq5K7g3dncTERK1evVpz5szJ2RYXF1dkQQEAAADSDSSrBw4cUJcuXfTKK69o4cKFkqTY2FiFh4dr586dRR4gAAAAiq9CJ6uvvvqqevbsqe+//15Wa/a3V6lSRWPGjNFrr71W5AECAADczuyyOPTlbAqdrO7du1ejR4+Wi4uLLH+Zvda3b1/t37+/SIMDAABA8VboCVaenp65ktQ/JScnX3U7AAAArs5ucfwTrOxOlq4V+u4EBwdr3rx5ubZdunRJU6dOVaNGjYosMAAAAKDQldVx48apf//++vjjj5Wenq6uXbsqNjZWXl5eevvtt29FjAAAALcvPpnOV6GT1dq1a2v9+vVat26dfv/9d3l4eKhGjRp64IEHVKpUqVsRIwAAAIqpG3ooQJkyZTRgwABJkt1u16+//qrMzMwiDQwAAAAodM/qzp071bFjR0mSzWZT//799eCDD6pt27bavn17kQcIAABwO7PL6tCXsyl0ZXXmzJnq3bu3JGnTpk06dOiQNmzYoJ9++klz5sxRixYtijxIAAAAFE+FTq8PHTqkgQMHSpK++eYbhYeHq0qVKurWrZsOHz5c1PEBAADcxiyyWxz7kpM9GKDQyaqLi4tcXFwkSdu3b1erVq0kZbcEZGRkFG10AAAAKNYK3QZQv359zZ07V+7u7kpKSsr52P/rr79W9erVizo+AAAAFGOFTlbHjx+vsWPHKikpSZMnT5anp6cuXLigqKgozZ49+xaECAAAcPty9BOsJJuDx7s5hU5W69atq/Xr1+fa5ufnpw0bNqhixYpFFhgAAABQ6FQ+MzNTH330Uc7Xmzdv1ogRI/Tvf/9b6enpRRocAADA7c4ui0NfzqbQyerMmTP13nvvSZJOnjypJ598Ul5eXtqxY4deffXVIg8QAAAAxVehk9UvvvhCCxYskCStW7dODRs21Kuvvqo5c+Zo06ZNRR4gAADA7cxusTr05WwKHXFSUpKqVasmSYqOjs55mlVAQIAuXLhQtNEBAACgWCt0slqmTBmdOHFC586d0+7du9W6dWtJ0qlTp1SyZMkiDxAAAADFV6FXA+jRo4d69+4tFxcXNWrUSEFBQUpJSVFUVJTatm17K2IEAAC4bWU/VQrXUuhkdfTo0brjjjuUlJSkBx54QJLk5uamatWqKSoqqsgDBAAAQPFV6GRVku6///5cX7u7u+ull15SVFSUpk+fXiSBAQAAFAfOuJyUI91Qsvrdd99p9+7dudZVPXnypP7zn/+QrAIAAKDIFDpZff/99zVt2jSVK1dO58+fV0BAgM6cOaOqVatq3LhxtyJGAAAAFFOFXg3g3//+txYuXKht27bJzc1N3377rTZt2qQaNWooNDT0VsQIAABwW7LL4vB1Vp2t7aDQyerZs2dzZv1b/jt7rWLFioqMjNT//d//FW10AAAAKNYKnayWLFlSZ8+elSR5eXkpNjZWknTHHXfo4MGDRRsdAADAbc4ui0NfzqbQyWrnzp3Vr18/JScnq3Hjxnr++ef15Zdfatq0afL3978VMQIAAKCYKnSyGhUVpTZt2sjDw0PPPvuszp49qzFjxmjNmjUaP378rYgRAADgtuXonlVnU+jVADw9PTVx4kRJUuXKlfXll1/q/Pnz8vPzk9XqfDcAAAAA5lWgZHXbtm0FPmGrVq1uOBgUL/e09FXXjn6q6O+uSylZ2r0/RV9tOZXrmJpVPBQ1tLIqBZTQg8MPKCPTblC0uN1ZLFKXFqXUJtRT5f1clZZu14Hf0/ThV5dyjqlXw1MjmvqqSoCbrqTZtGP/Fa34+pLSM3hfwhjPjwlWw/pl9FDE9zp9Ns3ocIBbokDJ6j/+8Q9ZLBbZ7fn/QLZYLPrll1+KJDDc3nrcU1aDegXo/VVn9P2eS6pY3l1P9g9UzSrVlHThV0lSeDs/DeoVoHMXMgyOFsVBny6l1bZxSS1Zl6RDx9MVUNZFA7v66LlBfpq5JE7e3t564kF/bd6ZqsWfXJSPl1XDe/nKx8uquSsSjQ4fxVDZsmVVuYqP0WGgCDjjpCdHKlCy+p///OdWx4Fipmfnstq0PVGfbDgvSTp1Ll0ffnZOT/YP1IFUT5Vwt6tvN3+9/NZx1a1ZUn27lTc4YtzOrFbp7voeWr8tRdE/p0qS4hOz9MmmZA3r5auK/u6qUMFP5y5k6r21FyVJJ89J76+7qDF9/VSpfLLizmYaeQkoZnx93FS5cmVt3HJa97YPNDoc4JYqULJaqVKlXF+fOnVKVqtVAQEBkqTDhw/L09Mzz3E36sSJE+rYsaPWr1+voKCgIjknzGXEC0dk+1ul/nxCdgXVarUqMzNDY6Yc1bkLGapbs6QRIaIYsdmksa+dy7P9z0+Tsmx2lSxZUvuOXs61f8+hNGVm2RUcVIJkFQ4V0ecOpaSkaMfOeJJVZ2eR4yc9OVkht9B354cfftD999+vn376KWdbdHS0HnjgAf34449FGpyjffDBB+rSpYtCQ0PVvXt3bdy40eiQblvJl7N0OdWWa1uzu0rrSlqWUlNTlWUTH//DUNUquKp7u9La9esVnTqXIbvdLtvfOqFsNulyqk0VyroYEySKpfYtyymkvq/++OMPo0MBHKLQqwG89tprevLJJ3XfffflbOvfv7/c3Nw0c+ZMrVixokgDdJSvvvpKs2bN0sKFC9WgQQOtWbNGY8aM0RdffKEqVaoYHd5tr2kDL3VpXUafbjijquVs1/8G4BZ55J7SurdFKVmt0sYfLuuDL5PkV1pKS0tTlQruuY71LW2Vt5eLPNydrEwBp1Xay1WRQ2vpg0+OqWHtDEluRocE3HKFrqz+9ttvGjhwYJ5lqh555BH99ttvhTpXbGysnnjiCYWGhqp9+/ZaunTpVY87fvy4IiIi1KxZMzVr1kxjx45VUlKSJMlms2natGlq1aqV7rrrLnXr1k1bt26VJKWmpioqKkotWrRQaGioevfurX379l11jCtXrmjs2LFq3Lix3Nzc9PDDD6tUqVLas2dPoa4JhdeysbeeG1ZF3+64qPXf5v0oFnCk9d8la+Jb8Vq4OlGN63pobD8/WSzZj5quWqGEHmhdSu5uUlkfq4b29FXyZZuy+P0KDvLU4CCdPJ2qr789df2D4SQc+/Sq7MlczvULdqErq97e3jp27Jhq1qyZa/uvv/6qkiUL11s4atQoNW3aVPPmzdOxY8f02GOPKSgoSNWqVct13MSJE1WpUiVt3bpVycnJioiI0Pz58zV+/Hh9/vnnio6O1tq1a+Xj46M1a9YoKipKmzdv1pIlSxQfH68NGzbI3d1dixcv1qRJk/TJJ5/kiaV79+65vk5KSlJKSkpOX25BuFjsqkhVsFA6tCirRx+oqG+/P68PPzulcr7Z9+/PPyXJq2T2Z68VytmUydJVheLtk250CE4pI106/Ie09LMriuxXUS1CSujChTgdPeOtB9uVUa+OpZWaZtNX0YkqXcpLmZkZ8udeF1haRdbkvhEN7yyj9i399fzU3SpfNjvZKOOT/WfF8la5unBfb4Sri5SZZXQUyE+hk9UePXpo6NCh6tevnypXriy73a4jR45o+fLl6t27d4HPc+DAAR08eFBLliyRp6en6tWrp7lz5141OVy0aJEsFovc3d3l5+en1q1ba9euXZKyk0pXV1d5enrKxcVFDz30kB588EFZrVYlJSXJzc1NHh4ecnV11YgRIzRixIjrxma32zVx4kQ1bNhQTZs2LfA1eXtJw3rxD1ZBlStXTlWrVlRcXJx8SpzR0If+t69Xp/9NVqlYMfunyOAH06+7fBr+Lt7oAJyCi4uLvL29denSJWVm/u+9Z7FYJFVUizttOnVK8nU7ppiYP+Tq6qqMjAxV85Uq+d+lzMvn1bddgmHxO512TJq8EdWqVVAJd6tm/bNxzrZGjbL/nPNyE126dKnQn3AiW3yCsYUmu8W5Kp2OVuhkddSoUbLZbJo/f74uXsxewqVMmTLq27evRo0aVeDzHD9+XF5eXvL19c3ZFhYWJil7NYC/2rdvn2bNmqWDBw8qIyNDWVlZCg4OliSFh4fr008/VZs2bdSyZUu1a9dO4eHhslqt6tu3ryIiItS2bVu1bt1anTp1UseOHfONKyMjQ+PHj9fhw4ev2ZZwLUnJ0gdfuV//QKhuzVKKjKiqjz4/pY3fJUjKvm/lfG3q1SlTqza6Kj4xu0rQtaOLugVKiz9xp7JaSN7lvI0OwSn4lnbRpCGV9dmWBH3zY1LO9qoV3NWokbT7N1fVLF9SB06UUfTP/1t4PeQOT4U0sOjd9W5KTStnROhO6cTB40aH4JTK+B6RV8nsSVVly1jUrVMJ7fnFTY/1qq2X39inU2dSde48DwYorCF9PIwOAddR6GTVxcVFkZGRioyMVEJCQk5ForCsVqtstuv/JnPx4kUNGTJEffr00eLFi+Xl5aXZs2crOjpakuTr66uPPvpIu3bt0jfffKM333xTH3zwgZYvX67KlStr/fr12rFjhzZt2qTJkydr7dq1evPNN6861pUrVzRixAilpqZq+fLlKlOmTKGuKctu0al4PoYpiElPBuqXI5e17ptL+jNRlaT0TJusVpsuJrsoNT174kBmVvZM6yvp7srItCsz067ky3xmUxBpbvzyVBDnLkpbdl1Wx6Y+OnXeooPH0lXW10U9O3kr8VKWomOuKKRbKT3YMUDJV5K051CaqlZwVc+OPlrzbbKOny30j9Ji7cQp2qVuxIlTV3L+XrmiVZ1b2XX4WPbPwl0xKTzB6gYZ3QJgl2S3O7ay6mxln5v6CVvYZO6vqlSpopSUFJ09e1bly2cv+L5x40Z5e3srMPB/a8YdPXpUKSkpioiIkJeXl6TsFoI/paWlyWazqVGjRmrUqJGGDx+uli1b6tdff1WNGjXk5uamsLAwhYWFadCgQerQoYMSEhLyxG632xUZGSlXV1e9//77KlGixA1fG/Ln7+emqoHZv8kum1Unz/6TJ0+qSYMsDXo49yoM702vLUmKOZii52Yeu+Vxonh5f91FJVzKUve2XvLr6qKLyTYd/CNdq/5zSRa7XefOxWvfH6V1f6vS6nuft85fzNKnm5O1Ycfl658cAHDDDCsH1KtXT/Xr19fs2bM1YcIExcXFacKECZoxY0au4wIDA2W1WrV79261aNFCH330keLj45WYmKjMzExNnTpViYmJevHFF1WmTBnt379fNptNgYGBGj16tKpXr67IyEiVLFlSu3fvlq+vr3x88j6ebt26dTp8+LDWrl1LonqLnbuQoQcG77/qvorlbBrWK13Ru9z18ddJVz0GuBUys6SP/5Osj/+TnGef/39/ZPznhyR9uOFKnv2AUQ4cuqiWXTcbHQZwSxn6mfWCBQsUFxensLAwDRs2TCNGjFCbNm1yHRMQEKCxY8fq+eefV/v27XXx4kXNnDlT6enp6tu3r55++mlZrVZ16dJFjRo10tSpUzVr1iz5+fnppZde0h9//KE2bdqoSZMmWrZsmebNm5dn2S1J+vjjjxUXF6emTZsqJCQk5zVx4kRH3Q4AAFAM2WV16MvZGNpoFRAQoCVLluTZXrlyZR08eDDn68GDB2vw4MG5jtm2bVvO32fPnn3V8wcGBurtt98uUCxXiwMAAADGuuH0+sSJE/r++++LMhYAAIBix/EPBXAuhU5WL1y4oH79+qlTp076xz/+IUk6d+6cHnjgAZ06xRM1AAAAUHQKnaxOmzZNbm5uWrlyZU7vZ+nSpVWnTh1Nnz69yAMEAABA8VXoZHXLli2aNm2aQkJC/vt0F8nDw0MTJ07M1UcKAACA63OGNoCtW7cqLCxMkZGRubavXr1adevWzTU5PSQkRD///LMkyWaz6fXXX1fHjh3VpEkTRUREKDY2tlBjF3qCVUZGRs66qH/l4eGhjIyMwp4OAAAAJrZ48WKtWrVK1apVu+r+Jk2a6F//+tdV9y1fvlzr1q3T4sWLFRAQoNdff10jR47Up59+mlP0vJ5CV1aDgoL05Zdf5tm+YsUK1axZs7CnAwAAKMYcW1XNrqwWrrpaokSJfJPV/KxYsUIDBw5UUFCQvLy8FBkZqSNHjmjv3r0FPkehK6uDBw/W008/rS+++EJZWVl66aWXtH//fv3888/XXEIKAAAAzql///757j916pQGDRqkffv2ydvbW6NHj1b37t115coVHT58WPXr18851svLS9WqVVNMTIzuuuuuAo1f6GT1nnvu0cKFC7V8+XJVrVpVu3fvVo0aNfT888+rQYMGhT0dAABAseaMy0n9yc/PT9WrV9fYsWN1xx13aMOGDXr22WdVvnx51axZU3a7Pc+TQ318fJSQkFDgMW7ooQAtWrRQixYtbuRbAQAAcJto166d2rVrl/N1eHi4NmzYoNWrV2vcuHGSJLvdflNjFChZffPNNzV69GhJ0muvvZbvsWPHjr2pgAAAAOC8KlWqpH379snX11dWq1WJiYm59icmJqps2bIFPl+BktX169fnJKufffbZNY+zWCwkqwAAAAVll+x2B7cB3FyhM5cPPvhAPj4+uv/++3O2HTlyRFWqVFGJEiVUq1Yt7d+/X02bNpUkJSUl6fjx44VqHS1QsvrX2f+bNm265nF/z5wBAABw+0pPT9dLL72kKlWqqG7duvrqq6+0ZcsWffTRR5KkPn36aNGiRWrTpo0CAgI0c+ZM1atXTyEhIQUe44Z6Vq8mJSVFXbp00Y4dO4rqlAAAALc9s0+w+jOxzMzMlCRt3LhRkhQTE6P+/fsrJSVFTz31lM6dO6fKlStr3rx5Cg4OliT17t1b586d0+OPP66UlBQ1a9ZMc+fOLdT4hU5WL1y4oKlTp2rPnj1KS0vL2Z6cnCw/P7/Cng4AAAAmFhMTc819FotFI0aM0IgRI665f/To0TntpDei0A8FmDJlin777TeFh4crISFBDz/8sGrVqqW6detq6dKlNxwIAAAA8HeFTlZ37Nihd955R2PHjpWrq6ueeuopvffee2rRokVOWRgAAADXZ5cc/gSrIpxf5RCFTlaTk5Pl7+8vKbu0+2f/Qv/+/bVkyZKijQ4AAADFWqGT1apVq+qLL76QJAUEBOi7776TlL3g68WLF4s2OgAAgNucoyurzqbQE6z+8Y9/aOzYsWrRooW6du2qyMhINWnSREeOHFFoaOitiBEAAADFVKGT1e7duysoKEi+vr4aOXKkrFardu3apbZt22r48OG3IkYAAIDblsMfCuBkCp2sfv311+rcubOk/y1XAAAAANwKhe5Zff7555Wenn4rYgEAAAByKXSyOnDgQM2cOVNJSUm3Ih4AAIBixSaLQ1/OptBtABs3btTp06e1bNkylS5dWm5ubrn2b9u2rciCAwAAQPFW4GTVbrfLYrGoU6dOtzIeAACAYsSI5aScq7pa4GS1Y8eOeuyxx9S/f395e3vfypgAAAAASYXoWW3VqpXmzp2rtm3bavLkyfrtt99uZVwAAABAwZPV//u//9PWrVs1btw47d69W926ddPjjz+ujRs3ym53tqfMAgAAmIPdbnHoy9kUajUALy8vPfbYY1q3bp2WLl0qf39/RUZGqmPHjlq8eDGPWwUAAECRKvRqAH9q0qSJmjRpogsXLmjt2rVavXq15s+fr927dxdlfAAAALc1x0+wci6FXmf1744ePaqDBw/q5MmT8vf3L4qYAAAAAEk3WFlNSUnRmjVr9OGHH+q3335Ty5YtNWPGDLVr166IwwMAAEBxVqhk9dChQ/r3v/+tdevWyW63q0ePHpo9e7aCgoJuVXwAAAC3LbskOeGkJ0cqcLLat29f7d69W1WqVNGTTz6pXr16ycvL61bGBgAAgGKuwMmqp6en3nrrLbVt21YWC78BAAAAFAVHT7BytiyuwMnqO++8cyvjAAAAAPK44aWrAAAAcPMcvVC/s1VWb3rpKgAAAOBWIVkFAACAadEGAAAAYCCbg8dztkqls8ULAACAYoTKKgAAgIEcPcHK2VBZBQAAgGmRrAIAAMC0aAMAAAAwjMXhT7BytpVWqawCAADAtKisAgAAGMQux0+wsjt0tJtHZRUAAACmRWUVAADAQI7vWXUuVFYBAABgWiSrAAAAMC3aAAAAAAxkc7YZTw5GZRUAAACmRWUVAADAQEywyh+VVQAAAJgWySoAAABMizYAAAAAo9gd/wQrZ3uEFZVVAAAAmBaVVQAAAAPZnazS6WhUVgEAAGBaVFYBAAAMYpdkc/DSVc5WyKWyCgAAANMiWQUAAIBp0QYAAABgIIcvXeVkqKwCAADAtKisAgAAGIilq/JHZRUAAACmRbIKAAAA06INAAAAwEB2B6+z6myorAIAAMC0qKwCAAAYyMYEq3xRWQUAAIBpUVkFAAAwjMWAhwI4V48slVUAAACYFskqAAAATIs2gCJky8pSwul4o8NwaqUsVkkllXQ+UQmnbUaH4/QSz5w3OoTbQnoFq9TOU3EHY3WC9+VNK1OhnNEhOD3vsjZJ6fIu66sydupON8Pqkmbo+HY5/glWzjafi3c4AAAATIvKKgAAgIFsTjbhydGorAIAAMC0SFYBAABgWrQBAAAAGMjRE6ycDZVVAAAAmBaVVQAAAKPYJTsTrPJFZRUAAACmRWUVAADAQDZH96w6WSGXyioAAABMi2QVAAAApkUbAAAAgIEcvnQVbQAAAABA0aCyCgAAYBC7HL90lbM9g4DKKgAAAEyLZBUAAACmRRsAAACAgRy+zqqTobIKAAAA06KyCgAAYCCHL13lZKisAgAAwLSorAIAABiIymr+qKwCAADAtEhWAQAAYFq0AQAAABjIZnfsE6ycDZVVAAAAmBaVVQAAAIPY7Y6fYOVsE7qorAIAAMC0SFYBAACQr61btyosLEyRkZF59q1fv15du3ZVaGioevbsqW3btuXss9lsev3119WxY0c1adJEERERio2NLdTYJKsAAAAG+rMVwFGvwlq8eLGmTJmiatWq5dn3yy+/KCoqSuPGjdP333+vgQMHatSoUTp9+rQkafny5Vq3bp0WLVqkb775RtWrV9fIkSNlL0QgJKsAAAC4phIlSmjVqlVXTVZXrlyptm3bqm3btipRooS6deum2rVra+3atZKkFStWaODAgQoKCpKXl5ciIyN15MgR7d27t8Djk6wCAAAYyGZ37Kuw+vfvr9KlS1913/79+1W/fv1c2+rXr6+YmBhduXJFhw8fzrXfy8tL1apVU0xMTIHHJ1kFAADADUlMTJSPj0+ubT4+PkpISNDFixdlt9uvub+gSFYBAABww67Xf1qY/tSrYZ1VAAAAA9md+AlWZcqUUWJiYq5tiYmJ8vPzk6+vr6xW61X3ly1btsBjUFkFAADADQkODta+fftybYuJiVHDhg1VokQJ1apVS/v378/Zl5SUpOPHj6tBgwYFHoNkFQAAwEBmX7oqP4888oiio6P17bffKi0tTatWrdKxY8fUrVs3SVKfPn20dOlSHTlyRMnJyZo5c6bq1aunkJCQAo9BGwAAAACu6c/EMjMzU5K0ceNGSdkV1Nq1a2vmzJl65ZVXFBcXpzvuuEMLFy6Uv7+/JKl37946d+6cHn/8caWkpKhZs2aaO3duocYnWQUAADDQjSwn5UjXW2aqc+fO6ty581X3WSwWjR49WqNHj77h8WkDAAAAgGmRrAIAAMC0aAMAAAAwiF1FP+mpIGM6EyqrAAAAMC0qqwAAAAZydGXV2VBZBQAAgGmRrAIAAMC0aAMAAAAwit2AdVadrO2AyioAAABMi8oqAACAgZhglT8qqwAAADAtKqsAAAAGstmMjsDcqKwCAADAtEhWAQAAYFq0AQAAABjELsdPsHK2+VxUVgEAAGBaVFYBAAAMxNJV+aOyCgAAANMiWQUAAIBp0QYAAABgIBttAPmisgoAAADTorIKAABgFLtkZ+2qfFFZBQAAgGlRWQUAADAQS1flj8oqAAAATItkFQAAAKZFGwAAAICBbDajIzA3KqsAAAAwLSqrAAAABrHL8ROsnG0+F5VVAAAAmBbJKgAAAEyLNgAAAAAD2Zztc3kHo7IKAAAA06KyCgAAYCCeYJU/KqsAAAAwLSqrAAAARrFLdkc3rTpZJZfKKgAAAEyLyipMwWKRHu1eWT3uraBKFT21+LUs/bgnQfPePaoz59KMDg/F1MrFTVUxwCPn6w8XZf959I8U9X9yp0FRobi5p6Wvunb0U0V/d11KydLu/Sn6asupXMfUrOKhqKGVVSmghB4cfkAZmU5WOgPyQbIKUxj1RJC6da6gdz44rHZNM/TpRmlQ7zs05+WG6jv8R2XygxcG+eCTWH373UmNeNxD8/91RafO2pSZxfsRjtHjnrIa1CtA7686o+/3XFLF8u56sn+galappqQLv0qSwtv5aVCvAJ27kGFwtLhRLF2VP1O2AZw4cUJ16tTRkSNHjA4FDuBildqFldPy1bHatuOc0tPTFfNLot7+9zFVquCpO6qXMjpEFGOpV2y6mJShzMxMXUzK0IXEDCVdyjQ6LBQTPTuX1abtifpkw3mdOpeuXfuT9eFn51SrRil5enqqhLtVfbv56+W3jmvrjxeNDhe4Jais/pfdbte8efP08ccfKzExUYGBgRo8eLB69OhhdGi3vSyb9FDEDklS5Yr/+/3pz6U8qKoCKK5GvHBEtr+ta3Q+IbuCarValZmZoTFTjurchQzVrVnSiBBxk+xy/NJVzvavKsnqfy1ZskRr1qzRO++8o2rVqmnDhg2KjIxU7dq1Vb9+faPDK3aqVymlgY9W1bYd8Tp8LMXocADAEMmXs/Jsa3ZXaV1Jy1JqaqqybK58/I/bnqFtALGxsXriiScUGhqq9u3ba+nSpVc97vjx44qIiFCzZs3UrFkzjR07VklJSZIkm82madOmqVWrVrrrrrvUrVs3bd26VZKUmpqqqKgotWjRQqGhoerdu7f27dt31THq1q2rWbNmqWbNmnJxcdG9996r0qVL6/Dhw7fm4nFVfXtWV2hoqF6ZEKof9yTo+VcOGB0Sirk6QV4aP7q+QkJCNG9aEz0zopZ8fdyMDgvFVNMGXurSuoy++PacbDab0eEADmFoZXXUqFFq2rSp5s2bp2PHjumxxx5TUFCQqlWrluu4iRMnqlKlStq6dauSk5MVERGh+fPna/z48fr8888VHR2ttWvXysfHR2vWrFFUVJQ2b96sJUuWKD4+Xhs2bJC7u7sWL16sSZMm6ZNPPskTS/PmzXP+fuXKFa1atUpWq1UtWrQo8PW4WHN/jI3C+/6nON1RJUk797nq/nuqqXbNUnrlzX083eMGWWQxOgSndjk1Q2V8XbXt+zi5Wezaud9N4fdUU6MG3np+yh5mXN+g0mVJsm5E42BvRTxaRTv2JuqHPWdUtZNUzvd/99KrZPb7sUI5G+1TheBqtSvTZuzPShszrPJlWLJ64MABHTx4UEuWLJGnp6fq1aunuXPnKiAgIM+xixYtksVikbu7u/z8/NS6dWvt2rVLkpSUlCRXV1d5enrKxcVFDz30kB588EFZrVYlJSXJzc1NHh4ecnV11YgRIzRixIh845o4caJWrVqlwMBAzZs3T/7+/gW+Jl9vi54dSs/Qzbpy5YruvEM6d+Z3Nbyznl6OqqSEhASjw0IxlHj+kCSp8Z3SlSvSnUFXdP7cH6pdu7b+75lAXbhwweAInVW60QE4HX9/f1WpUkXnzp2Tuy1WvTplb+/V6X+T/SpWzG4ZGPxguuz8hl8oF5KMjgD5MSxZPX78uLy8vOTr65uzLSwsTFL2agB/tW/fPs2aNUsHDx5URkaGsrKyFBwcLEkKDw/Xp59+qjZt2qhly5Zq166dwsPDZbVa1bdvX0VERKht27Zq3bq1OnXqpI4dO+Yb15QpUzRx4kR9/vnnGjZsmJYsWVLgntXEJLsWf5haiLsASSrt5argur46cOiiPNwzNeAhTy35OFUXElO1bL60c7+LVq27bHSYTonKatEoX86iAQ95aMnHV5R6JUOLX5O+323Rp1/y//uNKF3Wx+gQnErbpn4KDQ3Ux1+e1ldb4iW5q5yvTb06ZWrVRlfFJ2Z/ote1o4u6BUqLP3GnsloIfe81/pcnfrfIn2HJqtVqLVC/zcWLFzVkyBD16dNHixcvlpeXl2bPnq3o6GhJkq+vrz766CPt2rVL33zzjd5880198MEHWr58uSpXrqz169drx44d2rRpkyZPnqy1a9fqzTffzHdMDw8PPfTQQ1q/fr1WrVqlyZMnF+iasmzSiVN8vFVY5ctZNGZIPc1776i2bI+TJJ2Jt8u7dPaSVUeOXeG+3iCLhWT1RlWr7KnHe1XV0pXHdTY++8EUZ+PtqhyY/b7cd/CyTpzmfXkjfO20SxVUg7ql1Kd7oN5ZeUafbrygv081iU+06lR89rbky9n/v5+Ot9KiUghGtwDg+gxLVqtUqaKUlBSdPXtW5cuXlyRt3LhR3t7eCgwMzDnu6NGjSklJUUREhLy8vCRltxD8KS0tTTabTY0aNVKjRo00fPhwtWzZUr/++qtq1KghNzc3hYWFKSwsTIMGDVKHDh2UkJCgMmXK5Ipn2LBhat26tR577LGcbRaLRa6uLJhwq52NT9PnG09r4KNVZbVkyt09TXfWcdNjD9VU/IU0ffPdOaNDRDF0Nj5NdwX76I6a9bTik2Nyd7fr7rs81a9XTR39I0XbfjhvdIgoBob1qaBfj1zW5h8uytf7f/8eeXvZZLVmys3VkrPdo0R20urr7aqMTLsyM+1XXU0A5kNlNX+GZWL16tVT/fr1NXv2bE2YMEFxcXGaMGGCZsyYkeu4wMBAWa1W7d69Wy1atNBHH32k+Ph4JSYmKjMzU1OnTlViYqJefPFFlSlTRvv375fNZlNgYKBGjx6t6tWrKzIyUiVLltTu3bvl6+srH5+8H0E1atRIixYtUmhoqGrXrq0tW7Zo+/btioiIcNQtKdZmzDukc+fT1DO8ivzLllDFSuna9fNFLfzX70pO4YctHC/1ik2jnt+rfzxWXcMG1lIZHzdVqpKhbTsuaNG/jimLp1jhFvP3c1PVwOzH/S6bVSfP/pMnT6pJgywNerhKru3vTa8tSYo5mKLnZh675XECt5qhZcMFCxbo2WefVVhYmMqWLasRI0aoTZs2uXpWAwICNHbsWD3//POSpL59+2rmzJnq37+/+vbtq8WLF+uFF15Qly5dlJmZqWrVqmnWrFny8/PTSy+9pMmTJ6tNmzayWCyqVauW5s2bJ6s170dQERERysjI0JAhQ3Tp0iVVrlxZU6ZMKdRqALhxGZl2LV52TF/857ieHVpSry68zEf/MNzps2ma8vpBVa5g1TNDPTVjYSof/cNhzl3I0AOD9191X8VyNg3rla7oXe76+GtmB+H2ZmiyGhAQoCVLluTZXrlyZR08eDDn68GDB2vw4MG5jtm2bVvO32fPnn3V8wcGBurtt98uUCwuLi4aOXKkRo4cWaDjAQAAbppdeZ5S5ogxnQld7gAAADAtZg8BAAAYyE53Ub6orAIAAMC0SFYBAABgWrQBAAAAGMQuOfzxuE42v4rKKgAAAMyLyioAAICBCvD0+WKNyioAAABMi8oqAACAYewO71nN7lq1OHjMG0dlFQAAAKZFsgoAAADTog0AAADAQDZnW0vKwaisAgAAwLSorAIAABjFLtkdXVp1skoulVUAAACYFskqAAAATIs2AAAAAAM5fJlVJ0NlFQAAAKZFZRUAAMAgdkk2B0+wcrZCLpVVAAAAmBbJKgAAAEyLNgAAAAAD2ZlhlS8qqwAAADAtKqsAAABGsUt2m+PHdCZUVgEAAGBaVFYBAAAMZKNnNV9UVgEAAGBaJKsAAAAwLdoAAAAADMTSVfmjsgoAAADTorIKAABgELskm82xlVVnq+NSWQUAAIBpkawCAADAtGgDAAAAMBDzq/JHZRUAAACmRWUVAADAQHYHT7ByNlRWAQAAYFpUVgEAAIxit8vm6KZVu12SxbFj3gSSVQAAAOSrTp06cnNzk8XyvyT3kUce0aRJk7R9+3bNmjVLR48eVcWKFTV06FB169atyMYmWQUAAMB1ffnll6pcuXKubWfPntWIESM0YcIEde3aVTt37tTw4cNVo0YNhYSEFMm49KwCAAAYyG6zO/RVlNatW6fq1aurV69eKlGihMLCwtShQwetXLmyyMYgWQUAAMB1zZo1S+3atdPdd9+tSZMmKSUlRfv371f9+vVzHVe/fn3t27evyMYlWQUAADCQM1RW77rrLoWFhenrr7/WihUrtGfPHv3zn/9UYmKivL29cx3r6+urhISEorg1kuhZBQAAwHWsWLEi5+9BQUEaN26chg8frsaNG9/ysamsAgAAoFAqV66srKwsWa1WJSYm5tqXkJAgPz+/IhuLZBUAAMAgdkk2u2NfhW0EOHDggKZNm5Zr25EjR+Tu7q62bdvm6U/dt2+fGjZseHM35i9IVgEAAHBNZcuW1YoVK7Ro0SKlp6fr999/1xtvvKFHH31U3bt3V1xcnFauXKm0tDRt3rxZmzdv1iOPPFJk49OzCgAAYBS7inw5qYKMWRgBAQFatGiRZs2apbfeekvu7u568MEHFRkZqRIlSmjhwoWaMmWK/vnPf6pSpUqaMWOG6tatW2ThkqwCAAAgX02aNNGHH354zX2ffvrpLRubZBUAAMBAdruDK6tOhp5VAAAAmBbJKgAAAEyLNgAAAAAD2Rw9wcrJUFkFAACAaVFZBQAAMBATrPJHZRUAAACmRbIKAAAA06INAAAAwCB22R3+BCt7YR9hZTAqqwAAADAtKqsAAABGscvhlVUnK6xSWQUAAIB5UVkFAAAwkI2lq/JFZRUAAACmRbIKAAAA06INAAAAwEAOn2DlZKisAgAAwLSorAIAABjIzgSrfFFZBQAAgGmRrAIAAMC0aAMAAAAwkI0JVvmisgoAAADTorIKAABgELvd8UtXOdt8LiqrAAAAMC0qqwAAAAZi6ar8UVkFAACAaVFZLSIZGRkq423R5KdKGh2KU3N1yf5zSB8PZWYZGwvwpz/fl4P7lOB9WQSsLmlGh+D0XK3Zlbi+96Yr02YxOBrn5uNll81mdBTID8lqEbFYLHJ1tSowwM3oUG4L5csZHQGQF+9LmE05P6MjcH4ZGRmyuBqZ8Ntld3i27FxtBySrRSQ0NNToEAAAAG47JKsAAAAG4qEA+WOCFQAAAEyLZBUAAACmRRsAAACAgVhnNX9UVgEAAGBaVFYBAACMYpfsjp5g5WSFXCqrAAAAMC0qqwAAAAaxy/GVVScrrFJZxa1z4sQJ1alTR0eOHDE6FEAS70mYD+9J4PpIVnHbSUlJ0bhx4/gHAKbxwQcfqEuXLgoNDVX37t21ceNGo0NCMWa32zV37ly1b99eoaGhCg8P15o1a4wOC7gmklXcVs6cOaOePXvKxcXF6FAASdJXX32lWbNm6eWXX9YPP/ygfv36acyYMYqNjTU6NBRTS5Ys0Zo1a/TOO+/op59+0pNPPqnnnntOBw4cMDq0Ysoum93m0JezNQKQrOKmxcbG6oknnlBoaKjat2+vpUuXXvW448ePKyIiQs2aNVOzZs00duxYJSUlSZJsNpumTZumVq1a6a677lK3bt20detWSVJqaqqioqLUokULhYaGqnfv3tq3b99Vx0hISNAzzzyjJ5988tZcLJyCmd6TV65c0dixY9W4cWO5ubnp4YcfVqlSpbRnz55bcu0wJzO9J+vWratZs2apZs2acnFx0b333qvSpUvr8OHDt+bigZtEsoqbNmrUKAUFBSk6Olrz58/X7Nmz9d133+U5buLEiSpfvry2bt2qL774Qr///rvmz58vSfr8888VHR2ttWvXaufOnRowYICioqKUkZGhJUuWKD4+Xhs2bNCOHTvUunVrTZo06aqx1K1bV506dbql1wvzM9N7snv37urbt2/O10lJSUpJSVFAQMCtuXiYkpnek82bN1fDhg0lZf8ytWzZMlmtVrVo0eLW3QDky26zO/TlbFgNADflwIEDOnjwoJYsWSJPT0/Vq1dPc+fOveo/xIsWLZLFYpG7u7v8/PzUunVr7dq1S1L2P+Curq7y9PSUi4uLHnroIT344IOyWq1KSkqSm5ubPDw85OrqqhEjRmjEiBGOvlQ4CTO/J+12uyZOnKiGDRuqadOmRX7tMCezvicnTpyoVatWKTAwUPPmzZO/v/8tuX7gZlFZxU05fvy4vLy85Ovrm7MtLCxMQUFBeY7dt2+fBg4cqEaNGikkJERvv/220tPTJUnh4eFydXVVmzZtNGbMGK1Zs0ZZWVmSpL59++r3339X27ZtNX78eP3nP/9xyLXBOZn1PZmRkaFx48bp8OHDeuONN4rmYuEUzPqenDJlivbs2aORI0dq2LBh9KzCtEhWcVOsVqtsNtt1j7t48aKGDBmiRo0aacuWLYqJidGQIUNy9vv6+uqjjz7SwoULVaVKFb355pvq16+fMjMzVblyZa1fv14zZsyQl5eXJk+erKeeeupWXhacmBnfk1euXNHQoUN18uRJLV++XOXKlSuSa4VzMON78k8eHh566KGH1KBBA61ateqmrhM3yG5AG4CTdQLQBoCbUqVKFaWkpOjs2bMqX768JGnjxo3y9vZWYGBgznFHjx5VSkqKIiIi5OXlJUm5fotPS0uTzWZTo0aN1KhRIw0fPlwtW7bUr7/+qho1asjNzU1hYWEKCwvToEGD1KFDByUkJKhMmTKOvWCYntnek3a7XZGRkXJ1ddX777+vEiVKOOAuwEzM9p4cNmyYWrdurcceeyxnm8VikasrKQHMicoqbkq9evVUv359zZ49WykpKTp06JAmTJigK1eu5DouMDBQVqtVu3fv1uXLl/X+++8rPj5e8fHxyszM1NSpUxUVFaULFy7Ibrdr//79stlsCgwM1OjRozV9+nQlJyfLZrNp9+7d8vX1lY+Pj0FXDTMz23ty3bp1OR/9k6gWT2Z7TzZq1EiLFi3SgQMHlJmZqU2bNmn79u1q3769o24J/sZutzv05WxIVnHTFixYoLi4OIWFhWnYsGEaMWKE2rRpk+uYgIAAjR07Vs8//7zat2+vixcvaubMmUpPT1ffvn319NNPy2q1qkuXLmrUqJGmTp2qWbNmyc/PTy+99JL++OMPtWnTRk2aNNGyZcs0b948Wa15377z589XSEiI7r33XknZM7FDQkJyZtOieDDTe/Ljjz9WXFycmjZtqpCQkJzXxIkTHXU7YAJmek9GRETokUce0ZAhQ9S4cWPNmjVLU6ZMYTUAmJbF7owpNgAAgJOLiYnRmfgMzfmolEPHffKRFAWUc1NISIhDx71RNKgAAAAYqCAT8Ioz2gAAAABgWlRWAQAADGKXHP5UKWfr/6SyCgAAANOisgoAAGAYu+x2R/esOldtlcoqAAAATItkFQAAAKZFsgrAdOrUqaMPPvjA6DBumejoaDVs2FCff/55vscdOXJEderU0Y4dOyRJTzzxhJ599llHhAjAUezZE6wc+XKyLgB6VgE4zvnz5/X2229r06ZNOn36tNzd3VW9enV1795dffv2verTdm43Z86c0SuvvKL3339foaGhhfred999N9fXy5Yt0/333y8/P7+iDBEATIVkFYBDxMbGqk+fPqpXr57eeOMN1a5dW5cvX9bmzZv10ksvaceOHZozZ47RYd5yGRkZevfdd+Xv739T50lKStLLL7+sFi1akKwCTs7RS1c5m9u/jAHAFF588UX5+PjorbfeUt26dWW1WuXl5aXw8HDNnTtXZcuW1cWLF/N8X1ZWll5//XW1adNGISEhateunWbNmpXriS/Lly9Xly5ddNddd6lFixZ67rnnlJycLEk6ceKEhg8frmbNmik0NFQ9evTQxo0bc7731KlTGj16tFq1aqWGDRuqV69eio6OvuZ1rF69Wnfeeae2bdum++67TyEhIerSpYu+//77nGM6dOigOXPm6NFHH1WzZs0kZT+hZsGCBRo8eLA6deqkdu3aafbs2crKysr5vo0bN+r+++/PiePXX3/NNfbjjz+uyMhI/frrrwoLC1NWVpa6d++e0xpQ2GsBAGdAsgrglrtw4YK+++47DRw4UK6ueT/Qufvuu3OS2b9bsmSJ/v3vf+vdd99VTEyM3nzzTS1dulQrVqyQJO3Zs0fTpk3T66+/rj179mjNmjX6/ffftWjRIkn/S5K//fZb/fjjjxo0aJCeeeYZJSYmKj09XQMHDlSJEiW0bt06/fDDD3rggQc0ZMgQHTly5JrXk5mZqWXLlun999/X999/r7CwMA0bNiwnQZakVatWacyYMdq+fbskae7cufrwww81Y8YM7d69W3PnztXq1atzqsknT57U6NGjFR4erh9//FGvvvqq3nvvvauOX7duXb3zzjuSpE8//VSvvvrqDV8LAJgdySqAWy42NlZ2u121atUq9Pf2799fGzZs0B133CFJatCggerWrau9e/dKyv44XJI8PT0lSQEBAfrggw80duzYnP2urq5yd3eXq6urunfvrl27dsnX11dbtmzR8ePHNXnyZJUpU0YlSpTQwIEDVb16dX322Wf5xjVkyBAFBASoVKlSevLJJ5WamqotW7bk7K9fv75atGghq9Uqm82m5cuXKyIiQsHBwbJarQoODtaAAQO0Zs0aSdIXX3yhUqVKaejQoXJ3d1fNmjU1cODAAt+nm7kWAMay2W0OfTkbelYB3HIWi0WS5ObmVujvTU5O1vTp07Vt2zYlJiZKyu77DAwMlCQ1b95cnTp1Unh4uBo0aKDmzZvrgQceyElux4wZo3Hjxumbb75R8+bN1aZNG913331yd3fX0aNHZbPZFBYWlmtMu92uuLi4fOP6a+Lt5+cnLy8vnTp1Kmdb1apVc/5+4cIFJSYmavr06Xr11VdzjSNJ6enpOnXqlCpUqJCr8lyY5P5mrgUAzIxkFcAtV716dVmtVsXExOjOO+8s1PeOGTNGZ86c0eLFi1WrVi25uLiob9++Ofvd3d31+uuvKzY2Vlu3btW3336rRYsWafLkyerdu7fCwsL07bffaseOHYqOjtasWbO0YMECrVy5Uh4eHipZsqR2795d6GvKzMzM9bXdbs+1msFfE3MPDw9J0owZM3Tfffdd9XxpaWl5tv2ZzBbEzVwLAGMxwSp/tAEAuOW8vb3Vrl07LVy4UKmpqXn2//LLL+rUqZNiY2Pz7Nu5c6d69uypunXrysXFRSkpKfrtt99y9mdmZiopKUlVqlRR3759tWjRIg0ZMkTLly+XlF3VdHd3V+vWrRUVFaX169fr9OnTio6OVo0aNXT58uU8PZ1/ti3k59ixYzl/v3DhglJSUlSxYsWrHuvl5SV/f3/t378/1/b4+HilpKRIkipUqKDTp0/nSoL/PsEqPzdzLQBgZiSrABzihRdekCQ9+uij+umnn5SVlaXk5GStW7dOgwYNUsuWLVWlSpU831etWjXt3btX6enpio2N1XPPPafAwECdPn1adrtdixcvVt++fXOStMTERB06dEg1a9bU5cuX1blzZ7377rtKTU2VzWbTzz//rPT0dFWvXl0tW7ZU7dq19eKLL+rkyZPKzMzU559/rvvuu0+7du3K93oWLVqks2fPKiUlRW+88YZKlSql1q1bX/P4gQMH6oMPPtCWLVuUmZmpo0ePKiIiQtOnT5ckdezYUZcuXdK7776r9PR0HT58WEuXLr3m+f7s0T169KguXbp0U9cCwDh2u2S32Rz7crLfX2kDAOAQFSpU0Mcff6zFixdrwoQJOn36tDw8PFSrVi1NnDhRDzzwwFW/7//+7/80efJkNWnSRFWrVtWzzz6rtLQ0RUVFqV+/fnrvvfeUkJCggQMH6uLFi/L29lZYWJjGjx+vkiVLasGCBZo5c6bmzJkji8WiqlWravr06apdu7Yk6a233tK0adPUrVs3paWlKSgoSK+//roaN26c7/U8/PDDGjRokI4fP65KlSpp8eLFKlWq1DWPHzRokK5cuaIXX3xRZ8+elY+Pj7p166YxY8ZIyp7hP2vWLM2ZM0fz5s1TUFCQnnzySQ0bNuyq56tXr55atGihyMhItWzZUgsXLrzhawEAM7PY+XwIAAps9erVeu655/Tzzz+rRIkSRocDwInFxMTo1Nl0zXj3+scWpWeekCqWd1dISIhjB75BVFYBAAAMY5cTriblUPSsAgAAwLRIVgGgEHr27KmDBw/SAgCgyNjtNoe+nA3JKgAAAEyLZBUAAACmxQQrAAAAA9l4glW+qKwCAADAtKisAgAAGOW/T7By7Jgujh3vJlFZBQAAgGlRWQUAADCQnZ7VfFFZBQAAgGmRrAIAAMC0aAMAAAAwjN2Ap0o5V9sBlVUAAACYFpVVAAAAg9jl+AlWzlVXpbIKAAAAEyNZBQAAgGnRBgAAAGAghz/BysmQrAIAABikjI9Fk54q6fAxnQnJKgAAgAHc3d0lSZ4exo3tDCx2u93ZJoUBAACgmGCCFQAAAEyLZBUAAACmRbIKAAAA0yJZBQAAgGmRrAIAAMC0SFYBAABgWiSrAAAAMC2SVQAAAJjW/wO4wYsJJC3yfAAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                     ___Support Vector Machine (SVM)___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.94      0.97      0.96      1322\n",
            "     class 2       0.72      0.67      0.69       231\n",
            "     class 3       0.88      0.79      0.83       147\n",
            "\n",
            "    accuracy                           0.91      1700\n",
            "   macro avg       0.85      0.81      0.83      1700\n",
            "weighted avg       0.91      0.91      0.91      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1278   40    4]\n",
            " [  65  154   12]\n",
            " [  10   21  116]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.9105882352941177\n",
            "______________________________Test______________________________________\n",
            "               ___Support Vector Machine (SVM)___\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.93      0.94      0.93       333\n",
            "     class 2       0.63      0.61      0.62        64\n",
            "     class 3       0.75      0.72      0.74        29\n",
            "\n",
            "    accuracy                           0.87       426\n",
            "   macro avg       0.77      0.76      0.76       426\n",
            "weighted avg       0.87      0.87      0.87       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[312  18   3]\n",
            " [ 21  39   4]\n",
            " [  3   5  21]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.8732394366197183\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_all_test, Y_predec_model_5_without, 'coolwarm')  \n",
        "                        ###########################################\n",
        "                        ####___Support Vector Machine (SVM) ___####\n",
        "                        ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")              \n",
        "print(\"                     ___Support Vector Machine (SVM)___\")            \n",
        "print(classification_report(Y_all_train, Y_predec_model_5_without_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_train, Y_predec_model_5_without_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_SVM_without = accuracy_score(Y_all_train, Y_predec_model_5_without_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_SVM_without)\n",
        "print(\"______________________________Test______________________________________\")              \n",
        "print(\"               ___Support Vector Machine (SVM)___\")            \n",
        "print(classification_report(Y_all_test, Y_predec_model_5_without, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_test, Y_predec_model_5_without))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_SVM_without = accuracy_score(Y_all_test, Y_predec_model_5_without)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_SVM_without)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "ebed5549",
      "metadata": {
        "id": "ebed5549",
        "outputId": "c6d36c4a-4a8b-49a6-fc86-d52a0c0fa380",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "______________________________Train______________________________________\n",
            "                      ______Naives Bayes______\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       0.99      0.78      0.87      1322\n",
            "     class 2       0.41      0.92      0.57       231\n",
            "     class 3       0.63      0.63      0.63       147\n",
            "\n",
            "    accuracy                           0.78      1700\n",
            "   macro avg       0.68      0.78      0.69      1700\n",
            "weighted avg       0.88      0.78      0.81      1700\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[1026  253   43]\n",
            " [   7  213   11]\n",
            " [   4   50   93]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.7835294117647059\n",
            "______________________________Test______________________________________\n",
            "               ______Naives Bayes______\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "     class 1       1.00      0.75      0.86       333\n",
            "     class 2       0.43      0.92      0.59        64\n",
            "     class 3       0.42      0.55      0.48        29\n",
            "\n",
            "    accuracy                           0.77       426\n",
            "   macro avg       0.62      0.74      0.64       426\n",
            "weighted avg       0.87      0.77      0.79       426\n",
            "\n",
            "--------------------------------------------------------------------\n",
            "[[251  64  18]\n",
            " [  1  59   4]\n",
            " [  0  13  16]]\n",
            "--------------------------------------------------------------------\n",
            "Accuracy: 0.7652582159624414\n"
          ]
        }
      ],
      "source": [
        "plot_confusion_matrix(Y_all_test, Y_predec_model_6_without, 'coolwarm')  \n",
        "                        ###########################################\n",
        "                        ##########_______naive bayes _______#######\n",
        "                        ###########################################\n",
        "\n",
        "print(\"______________________________Train______________________________________\")              \n",
        "print(\"                      ______Naives Bayes______\")            \n",
        "print(classification_report(Y_all_train, Y_predec_model_6_without_train, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_train, Y_predec_model_6_without_train))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_naive_Bayes_without = accuracy_score(Y_all_train, Y_predec_model_6_without_train)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_naive_Bayes_without)\n",
        "print(\"______________________________Test______________________________________\")              \n",
        "print(\"               ______Naives Bayes______\")            \n",
        "print(classification_report(Y_all_test, Y_predec_model_6_without, target_names=target_names))\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(confusion_matrix(Y_all_test, Y_predec_model_6_without))\n",
        "# Calculer l'exactitude du modèle\n",
        "accuracy_naive_Bayes_without = accuracy_score(Y_all_test, Y_predec_model_6_without)\n",
        "print(\"--------------------------------------------------------------------\")\n",
        "print(\"Accuracy:\", accuracy_naive_Bayes_without)"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "8d518bb8",
      "metadata": {
        "id": "8d518bb8"
      },
      "source": [
        "## V.Interprétation des résultats :\n",
        "    Après avoir évalué les performances du modèle, on va interpréter les résultats et proposer des améliorations pour améliorer les performances.\n",
        "    -----------------------------------------------------------------------------------------------------------------------\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0VYdLzv7OziB"
      },
      "source": [
        "## V.le meilleur modèle  :\n",
        "    Après avoir évalué et intérprété chacun du modèle, on va comparer les résultats .\n",
        "    -----------------------------------------------------------------------------------------------------------------------\n"
      ],
      "id": "0VYdLzv7OziB"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "5469a4a8",
      "metadata": {
        "id": "5469a4a8"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "07a12ef2",
      "metadata": {
        "id": "07a12ef2"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "bdd23f07",
      "metadata": {
        "id": "bdd23f07"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "915ff2fc",
      "metadata": {
        "id": "915ff2fc"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "21438c00",
      "metadata": {
        "id": "21438c00"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "f3aa38eb",
      "metadata": {
        "id": "f3aa38eb"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "42b5003c",
      "metadata": {
        "id": "42b5003c"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.12"
    },
    "colab": {
      "provenance": [],
      "collapsed_sections": [
        "b2363adb"
      ]
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}