Coding_Challenge_TrainingTweedie_Regression.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mqF03oHgCQTL",
        "outputId": "c40f7a9b-0ec0-4120-d36f-4072e4bfd738"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting arff\n",
            "  Downloading arff-0.9.tar.gz (4.7 kB)\n",
            "  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Building wheels for collected packages: arff\n",
            "  Building wheel for arff (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for arff: filename=arff-0.9-py3-none-any.whl size=4950 sha256=439aaea544ad7eed059105741c15ecd512dc319c77b51505fe304a96f9354ea0\n",
            "  Stored in directory: /root/.cache/pip/wheels/0c/39/12/4d8ff491018bce2dd9cc6034298f27399c7f4fd4063187f7be\n",
            "Successfully built arff\n",
            "Installing collected packages: arff\n",
            "Successfully installed arff-0.9\n",
            "Collecting optuna\n",
            "  Downloading optuna-3.6.1-py3-none-any.whl (380 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m380.1/380.1 kB\u001b[0m \u001b[31m7.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting alembic>=1.5.0 (from optuna)\n",
            "  Downloading alembic-1.13.1-py3-none-any.whl (233 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m233.4/233.4 kB\u001b[0m \u001b[31m28.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting colorlog (from optuna)\n",
            "  Downloading colorlog-6.8.2-py3-none-any.whl (11 kB)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from optuna) (1.25.2)\n",
            "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from optuna) (24.0)\n",
            "Requirement already satisfied: sqlalchemy>=1.3.0 in /usr/local/lib/python3.10/dist-packages (from optuna) (2.0.29)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from optuna) (4.66.2)\n",
            "Requirement already satisfied: PyYAML in /usr/local/lib/python3.10/dist-packages (from optuna) (6.0.1)\n",
            "Collecting Mako (from alembic>=1.5.0->optuna)\n",
            "  Downloading Mako-1.3.3-py3-none-any.whl (78 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m78.8/78.8 kB\u001b[0m \u001b[31m12.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: typing-extensions>=4 in /usr/local/lib/python3.10/dist-packages (from alembic>=1.5.0->optuna) (4.11.0)\n",
            "Requirement already satisfied: greenlet!=0.4.17 in /usr/local/lib/python3.10/dist-packages (from sqlalchemy>=1.3.0->optuna) (3.0.3)\n",
            "Requirement already satisfied: MarkupSafe>=0.9.2 in /usr/local/lib/python3.10/dist-packages (from Mako->alembic>=1.5.0->optuna) (2.1.5)\n",
            "Installing collected packages: Mako, colorlog, alembic, optuna\n",
            "Successfully installed Mako-1.3.3 alembic-1.13.1 colorlog-6.8.2 optuna-3.6.1\n"
          ]
        }
      ],
      "source": [
        "!pip install arff\n",
        "!pip install optuna"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import arff\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn import preprocessing\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.linear_model import LinearRegression, TweedieRegressor\n",
        "import optuna\n",
        "from sklearn.model_selection import cross_val_score, KFold\n",
        "from sklearn.metrics import mean_squared_error, make_scorer, mean_absolute_error\n",
        "import seaborn as sns"
      ],
      "metadata": {
        "id": "FV8lK1DQDFYh"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Load Data"
      ],
      "metadata": {
        "id": "hiiZ4cFzDNS9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "data_freq = arff.load('freMTPL2freq.arff')\n",
        "df_freq = pd.DataFrame(data_freq, columns=[\"IDpol\", \"ClaimNb\", \"Exposure\", \"Area\", \"VehPower\", \"VehAge\", \"DrivAge\", \"BonusMalus\", \"VehBrand\", \"VehGas\", \"Density\", \"Region\"])\n",
        "data_sev = arff.load('freMTPL2sev.arff')\n",
        "df_sev = pd.DataFrame(data_sev, columns=[\"IDpol\", \"ClaimAmount\"])\n",
        "del data_freq, data_sev"
      ],
      "metadata": {
        "id": "Bt30bC8jDKyv"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# First sum up the \"ClaimAmount\" with duplicate entries of \"IDpol\".\n",
        "df_sev = df_sev.groupby(\"IDpol\", as_index=False).agg({'ClaimAmount': 'sum'})\n",
        "\n",
        "# There are some nan features in df_freq with some \"IDpol\" in df_freq, which is directly cleaned,\n",
        "merged_df = pd.merge(df_freq, df_sev, on=\"IDpol\", how=\"outer\")\n",
        "merged_df[\"ClaimAmount\"] = merged_df[\"ClaimAmount\"].fillna(0)\n",
        "merged_df = merged_df.dropna(how=\"any\")\n",
        "merged_df = merged_df.drop(['IDpol','ClaimNb'], axis = 1)"
      ],
      "metadata": {
        "id": "NHlmy-_-DPKX"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Hyper-parameters\n",
        "random_seed = 35\n",
        "np.random.seed(random_seed)\n",
        "input_standardization = True\n",
        "output_standardization = False\n",
        "feature_visualization = True\n"
      ],
      "metadata": {
        "id": "HvikXK-mDYCL"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Feature Preprocessing"
      ],
      "metadata": {
        "id": "Gdc8EyzQD4VZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the column transformer for handling different types of data\n",
        "preprocessor = ColumnTransformer(\n",
        "    transformers=[\n",
        "        ('num', preprocessing.MinMaxScaler(), [col for col in merged_df.columns if merged_df[col].dtype in [np.float64, np.float32] and col not in ['ClaimAmount', 'Exposure', 'IDpol', 'ClaimNb']]),\n",
        "        ('cat', preprocessing.OneHotEncoder(handle_unknown='ignore'), [col for col in merged_df.columns if merged_df[col].dtype == 'object']),\n",
        "    ],\n",
        "    remainder='drop'\n",
        ")\n",
        "\n",
        "# Exclude 'ClaimAmount' and 'Exposure' from any transformation if output_standardization is applied\n",
        "if output_standardization:\n",
        "    output_scaler = preprocessing.MinMaxScaler()\n",
        "claim_amount = merged_df.pop('ClaimAmount') / merged_df.pop('Exposure')  # Separate and transform this column separately\n",
        "\n",
        "# Create a processing pipeline\n",
        "pipeline = Pipeline(steps=[\n",
        "    ('preprocessor', preprocessor),\n",
        "    # ('output_scaler', output_scaler) if output_standardization else ('dummy', 'passthrough')\n",
        "])\n",
        "\n",
        "# Fit and transform the training data\n",
        "X = pipeline.fit_transform(merged_df)\n",
        "y = claim_amount\n",
        "\n",
        "\n",
        "# If there's output standardization, fit and transform 'ClaimAmount' specifically\n",
        "if output_standardization:\n",
        "    y = output_scaler.fit_transform(y.values.reshape(-1, 1))\n",
        "\n",
        "\n",
        "# Convert sparse Matrice representation of one-hot encoding to full representation\n",
        "X = X.toarray()\n",
        "print(X[0], X[0].shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PHEcLtamDvCm",
        "outputId": "1c3539d7-0137-4817-f68c-79ec4c398498"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0.09090909 0.         0.45121951 0.         0.04503871 0.\n",
            " 0.         0.         1.         0.         0.         0.\n",
            " 0.         0.         1.         0.         0.         0.\n",
            " 0.         0.         0.         0.         0.         1.\n",
            " 0.         0.         0.         0.         0.         0.\n",
            " 0.         0.         0.         0.         0.         0.\n",
            " 0.         0.         0.         0.         0.         1.\n",
            " 0.         0.         0.         0.        ] (46,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Define Optuna Objective Function for hyperparameter optimization"
      ],
      "metadata": {
        "id": "u-cSh-vsEBhR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def objective(trial, X, y):\n",
        "    # Suggest values for the hyperparameters\n",
        "    power = trial.suggest_float('power', 1, 2)\n",
        "    alpha = trial.suggest_float('alpha', 1e-3, 10.0, log=True)  # Regularization strength\n",
        "    solver = trial.suggest_categorical('solver', ['lbfgs', 'newton-cholesky'])\n",
        "    # Create a model with suggested hyperparameters\n",
        "    model = TweedieRegressor(power=power, alpha=alpha, solver=solver, max_iter= 300)\n",
        "    # Perform 5-fold cross-validation\n",
        "    kf = KFold(n_splits=5, shuffle=True, random_state=random_seed)\n",
        "    scores = cross_val_score(model, X, y, cv=kf, scoring=make_scorer(mean_squared_error))\n",
        "    # Compute RMSE from scores\n",
        "    rmse = (scores.mean()) ** 0.5\n",
        "    return rmse\n",
        "\n",
        "def objective_wrapper(trial):\n",
        "    return objective(trial, X, y)"
      ],
      "metadata": {
        "id": "jarwex7VD-Ky"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Call Optuna for Hyper-parameter Optimization\n",
        "Assign the incumbent setting to the model"
      ],
      "metadata": {
        "id": "PoO85PXgEMqj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "study = optuna.create_study(direction='minimize')\n",
        "study.optimize(objective_wrapper, n_trials=10)\n",
        "\n",
        "# Best trial result\n",
        "print(f\"Best trial: {study.best_trial.value}\")\n",
        "print(f\"Best parameters: {study.best_trial.params}\")\n",
        "best_params = study.best_params.copy()\n",
        "\n",
        "model = TweedieRegressor(**best_params)\n",
        "model.fit(X,y)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 268
        },
        "id": "n3lgX1FgEGZC",
        "outputId": "859b5d1d-fef1-4a35-8080-ac6d5c6583fc"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[I 2024-05-01 21:36:01,811] A new study created in memory with name: no-name-2539cfcd-eb11-48e8-8c4d-0e1af012bf80\n",
            "[I 2024-05-01 21:37:35,761] Trial 0 finished with value: 36820.537207231544 and parameters: {'power': 1.0499725434550211, 'alpha': 0.04515810515320362, 'solver': 'lbfgs'}. Best is trial 0 with value: 36820.537207231544.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best trial: 36820.537207231544\n",
            "Best parameters: {'power': 1.0499725434550211, 'alpha': 0.04515810515320362, 'solver': 'lbfgs'}\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_glm/glm.py:284: 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",
            "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TweedieRegressor(alpha=0.04515810515320362, power=1.0499725434550211)"
            ],
            "text/html": [
              "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>TweedieRegressor(alpha=0.04515810515320362, power=1.0499725434550211)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TweedieRegressor</label><div class=\"sk-toggleable__content\"><pre>TweedieRegressor(alpha=0.04515810515320362, power=1.0499725434550211)</pre></div></div></div></div></div>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Visualize the feature weight"
      ],
      "metadata": {
        "id": "A3Y__6bAE0le"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Get feature names from the column transformer\n",
        "feature_names = preprocessor.get_feature_names_out()\n",
        "# print(\"New Feature Names:\", feature_names)\n",
        "\n",
        "# Plot feature importance\n",
        "plt.figure()\n",
        "# plt.bar(range(len(feature_importances)), feature_importances, alpha=0.7)\n",
        "\n",
        "# Step to categorize and average gradients\n",
        "feature_groups = {}\n",
        "for name, weight in zip(feature_names, model.coef_):\n",
        "    # Parse the feature name to find the base feature (before one-hot encoding)\n",
        "    if '__' in name:\n",
        "        base_name = name.split('__')[1].split('_')[0]\n",
        "        if base_name not in feature_groups:\n",
        "            feature_groups[base_name] = []\n",
        "        feature_groups[base_name].append(abs(weight))\n",
        "\n",
        "# Calculating the mean gradient for each original feature\n",
        "mean_weights = {feature: np.mean(weight)  for feature, weight in feature_groups.items()}\n",
        "\n",
        "# Sorting for better visualization\n",
        "sorted_features = sorted(mean_weights, key=mean_weights.get, reverse=True)\n",
        "sorted_importances = [mean_weights[feature] for feature in sorted_features]\n",
        "\n",
        "\n",
        "plt.barh(sorted_features, sorted_importances, color='skyblue')\n",
        "plt.ylabel('Feature Index')\n",
        "plt.xlabel('Importance')\n",
        "plt.title('Feature Importance from Tweedie Regression')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "PYpsKjEeFWeU",
        "outputId": "87a5178b-d685-43b9-ea5e-9451486cce92"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}