kannada-mnist.ipynb

{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport matplotlib.pyplot as plt\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 5GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","execution_count":41,"outputs":[{"output_type":"stream","text":"/kaggle/input/kannada-mnist/Kannada_MNIST_datataset_paper.zip\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/kannada_mnist_arxiv_submission.pdf\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Dig_MNIST/y_dig_MNIST.npz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Dig_MNIST/X_dig_MNIST.npz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST/X_kannada_MNIST_train.npz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST/y_kannada_MNIST_test.npz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST/X_kannada_MNIST_test.npz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST/y_kannada_MNIST_train.npz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_Ubyte_gz/Dig_MNIST/X_dig_MNIST-idx3-ubyte.gz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_Ubyte_gz/Dig_MNIST/y_dig_MNIST-idx1-ubyte.gz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_Ubyte_gz/Kannada_MNIST/X_kannada_MNIST_train-idx3-ubyte.gz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_Ubyte_gz/Kannada_MNIST/y_kannada_MNIST_test-idx1-ubyte.gz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_Ubyte_gz/Kannada_MNIST/y_kannada_MNIST_train-idx1-ubyte.gz\n/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_Ubyte_gz/Kannada_MNIST/X_kannada_MNIST_test-idx3-ubyte.gz\n/kaggle/input/kannada-mnist/Kannada_MNIST_datataset_paper/kannada_mnist_arxiv_submission.pdf\n/kaggle/input/kannada-mnist/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Dig_MNIST/y_dig_MNIST.npz\n/kaggle/input/kannada-mnist/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Dig_MNIST/X_dig_MNIST.npz\n/kaggle/input/kannada-mnist/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST/X_kannada_MNIST_train.npz\n/kaggle/input/kannada-mnist/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST/y_kannada_MNIST_test.npz\n/kaggle/input/kannada-mnist/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST/X_kannada_MNIST_test.npz\n/kaggle/input/kannada-mnist/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST/y_kannada_MNIST_train.npz\n","name":"stdout"}]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"dir = '/kaggle/input/kannada-mnist/kannada_mnist_datataset_paper/Kannada_MNIST_datataset_paper/Kannada_MNIST_npz/Kannada_MNIST'\n\nX_trai = np.load(os.path.join(dir,'X_kannada_MNIST_train.npz'))['arr_0']\nX_tes = np.load(os.path.join(dir,'X_kannada_MNIST_test.npz'))['arr_0']\ny_train = np.load(os.path.join(dir,'y_kannada_MNIST_train.npz'))['arr_0']\ny_test = np.load(os.path.join(dir,'y_kannada_MNIST_test.npz'))['arr_0']\n\nprint(X_trai.shape, X_tes.shape)\nprint(y_train.shape, y_test.shape)","execution_count":24,"outputs":[{"output_type":"stream","text":"(60000, 28, 28) (10000, 28, 28)\n(60000,) (10000,)\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_train = pd.DataFrame(X_trai.reshape(X_trai.shape[0], 784))\nX_test = pd.DataFrame(X_tes.reshape(X_tes.shape[0], 784))\ny_train=pd.DataFrame(y_train)\ny_test=pd.DataFrame(y_test)\ny_train.head()","execution_count":45,"outputs":[{"output_type":"execute_result","execution_count":45,"data":{"text/plain":"     0    1    2    3    4    5    6    7    8    9\n0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0\n1  0.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0\n2  0.0  0.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0\n3  0.0  0.0  0.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0\n4  0.0  0.0  0.0  0.0  1.0  0.0  0.0  0.0  0.0  0.0","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>0</th>\n      <th>1</th>\n      <th>2</th>\n      <th>3</th>\n      <th>4</th>\n      <th>5</th>\n      <th>6</th>\n      <th>7</th>\n      <th>8</th>\n      <th>9</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>1.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0.0</td>\n      <td>1.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>1.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>1.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>1.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"import seaborn as sns\ng=sns.countplot(y_train[0])\nprint(y_train[0].value_counts())\n\n","execution_count":36,"outputs":[{"output_type":"stream","text":"9    6000\n8    6000\n7    6000\n6    6000\n5    6000\n4    6000\n3    6000\n2    6000\n1    6000\n0    6000\nName: 0, dtype: int64\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#normalizing data\n#greyscale normalization\nX_train=X_train.astype('float32')/255.0\nX_test=X_test.astype('float32')/255.0\n\nfrom keras.utils.np_utils import to_categorical\ny_train=to_categorical(y_train, num_classes=10)\n","execution_count":39,"outputs":[{"output_type":"stream","text":"Using TensorFlow backend.\n","name":"stderr"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_train=X_train.values.reshape(-1,28,28,1)\nX_test=X_test.values.reshape(-1,28,28,1)","execution_count":47,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.imshow(X_train[2][:,:,0])","execution_count":48,"outputs":[{"output_type":"execute_result","execution_count":48,"data":{"text/plain":"<matplotlib.image.AxesImage at 0x7fa60d363390>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"from keras.models import Sequential\nfrom keras.layers import MaxPooling2D, Flatten, Dense, Dropout, BatchNormalization\nfrom keras.layers.convolutional import Conv2D\nimport tensorflow as tf\n\nmodel = tf.keras.models.Sequential([\n    tf.keras.layers.Conv2D(64, (3,3), padding='same', input_shape=(28, 28, 1)),\n    tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer=\"uniform\"),\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n    tf.keras.layers.Conv2D(64,  (3,3), padding='same'),\n    tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer=\"uniform\"),\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n    tf.keras.layers.Conv2D(64,  (3,3), padding='same'),\n    tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer=\"uniform\"),\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n\n    tf.keras.layers.MaxPooling2D(2, 2),\n    tf.keras.layers.Dropout(0.25),\n    \n    tf.keras.layers.Conv2D(128, (3,3), padding='same'),\n    tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer=\"uniform\"),\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n    tf.keras.layers.Conv2D(128, (3,3), padding='same'),\n    tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer=\"uniform\"),\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n    tf.keras.layers.Conv2D(128, (3,3), padding='same'),\n    tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer=\"uniform\"),\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n    \n    tf.keras.layers.MaxPooling2D(2,2),\n    tf.keras.layers.Dropout(0.25),    \n    \n    tf.keras.layers.Conv2D(256, (3,3), padding='same'),\n    tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer=\"uniform\"),\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n    tf.keras.layers.Conv2D(256, (3,3), padding='same'),\n    tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer=\"uniform\"),##\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n\n    tf.keras.layers.MaxPooling2D(2,2),\n    tf.keras.layers.Dropout(0.25),\n    \n    \n    tf.keras.layers.Flatten(),\n    tf.keras.layers.Dense(256),\n    tf.keras.layers.LeakyReLU(alpha=0.1),\n \n    tf.keras.layers.BatchNormalization(),\n    tf.keras.layers.Dense(10, activation='softmax')\n])\n","execution_count":106,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.summary()\n","execution_count":107,"outputs":[{"output_type":"stream","text":"Model: \"sequential_16\"\n_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\nconv2d_56 (Conv2D)           (None, 28, 28, 64)        640       \n_________________________________________________________________\nbatch_normalization_50 (Batc (None, 28, 28, 64)        256       \n_________________________________________________________________\nleaky_re_lu (LeakyReLU)      (None, 28, 28, 64)        0         \n_________________________________________________________________\nconv2d_57 (Conv2D)           (None, 28, 28, 64)        36928     \n_________________________________________________________________\nbatch_normalization_51 (Batc (None, 28, 28, 64)        256       \n_________________________________________________________________\nleaky_re_lu_1 (LeakyReLU)    (None, 28, 28, 64)        0         \n_________________________________________________________________\nconv2d_58 (Conv2D)           (None, 28, 28, 64)        36928     \n_________________________________________________________________\nbatch_normalization_52 (Batc (None, 28, 28, 64)        256       \n_________________________________________________________________\nleaky_re_lu_2 (LeakyReLU)    (None, 28, 28, 64)        0         \n_________________________________________________________________\nmax_pooling2d_20 (MaxPooling (None, 14, 14, 64)        0         \n_________________________________________________________________\ndropout_28 (Dropout)         (None, 14, 14, 64)        0         \n_________________________________________________________________\nconv2d_59 (Conv2D)           (None, 14, 14, 128)       73856     \n_________________________________________________________________\nbatch_normalization_53 (Batc (None, 14, 14, 128)       512       \n_________________________________________________________________\nleaky_re_lu_3 (LeakyReLU)    (None, 14, 14, 128)       0         \n_________________________________________________________________\nconv2d_60 (Conv2D)           (None, 14, 14, 128)       147584    \n_________________________________________________________________\nbatch_normalization_54 (Batc (None, 14, 14, 128)       512       \n_________________________________________________________________\nleaky_re_lu_4 (LeakyReLU)    (None, 14, 14, 128)       0         \n_________________________________________________________________\nconv2d_61 (Conv2D)           (None, 14, 14, 128)       147584    \n_________________________________________________________________\nbatch_normalization_55 (Batc (None, 14, 14, 128)       512       \n_________________________________________________________________\nleaky_re_lu_5 (LeakyReLU)    (None, 14, 14, 128)       0         \n_________________________________________________________________\nmax_pooling2d_21 (MaxPooling (None, 7, 7, 128)         0         \n_________________________________________________________________\ndropout_29 (Dropout)         (None, 7, 7, 128)         0         \n_________________________________________________________________\nconv2d_62 (Conv2D)           (None, 7, 7, 256)         295168    \n_________________________________________________________________\nbatch_normalization_56 (Batc (None, 7, 7, 256)         1024      \n_________________________________________________________________\nleaky_re_lu_6 (LeakyReLU)    (None, 7, 7, 256)         0         \n_________________________________________________________________\nconv2d_63 (Conv2D)           (None, 7, 7, 256)         590080    \n_________________________________________________________________\nbatch_normalization_57 (Batc (None, 7, 7, 256)         1024      \n_________________________________________________________________\nleaky_re_lu_7 (LeakyReLU)    (None, 7, 7, 256)         0         \n_________________________________________________________________\nmax_pooling2d_22 (MaxPooling (None, 3, 3, 256)         0         \n_________________________________________________________________\ndropout_30 (Dropout)         (None, 3, 3, 256)         0         \n_________________________________________________________________\nflatten_7 (Flatten)          (None, 2304)              0         \n_________________________________________________________________\ndense_17 (Dense)             (None, 256)               590080    \n_________________________________________________________________\nleaky_re_lu_8 (LeakyReLU)    (None, 256)               0         \n_________________________________________________________________\nbatch_normalization_58 (Batc (None, 256)               1024      \n_________________________________________________________________\ndense_18 (Dense)             (None, 10)                2570      \n=================================================================\nTotal params: 1,926,794\nTrainable params: 1,924,106\nNon-trainable params: 2,688\n_________________________________________________________________\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"from keras.optimizers import RMSprop\noptimizer = RMSprop(lr=0.001, rho=0.9, epsilon=1e-08, decay=0.0)","execution_count":108,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.compile(optimizer=optimizer, loss='categorical_crossentropy',\n                   metrics=['accuracy'])","execution_count":109,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from keras.preprocessing.image import ImageDataGenerator\ndatagen = ImageDataGenerator(zoom_range = 0.1,\n                            height_shift_range = 0.1,\n                            width_shift_range = 0.1,\n                            rotation_range = 10)","execution_count":110,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from tensorflow.keras.callbacks import LearningRateScheduler\nannealer = LearningRateScheduler(lambda x: 1e-3 * 0.9 ** x)","execution_count":111,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.fit_generator(datagen.flow(X_train, y_train, batch_size=16),\n                           steps_per_epoch=500,\n                           epochs=40, #Increase this when not on Kaggle kernel\n                           verbose=1,  #1 for ETA, 0 for silent\n                           callbacks=[annealer])","execution_count":113,"outputs":[{"output_type":"stream","text":"Epoch 1/40\n500/500 [==============================] - 139s 277ms/step - loss: 0.0847 - accuracy: 0.9724 - lr: 0.0010\nEpoch 2/40\n500/500 [==============================] - 126s 252ms/step - loss: 0.0736 - accuracy: 0.9771 - lr: 9.0000e-04\nEpoch 3/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0670 - accuracy: 0.9810 - lr: 8.1000e-04\nEpoch 4/40\n500/500 [==============================] - 127s 253ms/step - loss: 0.0568 - accuracy: 0.9830 - lr: 7.2900e-04\nEpoch 5/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0634 - accuracy: 0.9825 - lr: 6.5610e-04\nEpoch 6/40\n500/500 [==============================] - 127s 255ms/step - loss: 0.0571 - accuracy: 0.9846 - lr: 5.9049e-04\nEpoch 7/40\n500/500 [==============================] - 127s 253ms/step - loss: 0.0530 - accuracy: 0.9844 - lr: 5.3144e-04\nEpoch 8/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0434 - accuracy: 0.9887 - lr: 4.7830e-04\nEpoch 9/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0456 - accuracy: 0.9869 - lr: 4.3047e-04\nEpoch 10/40\n500/500 [==============================] - 132s 265ms/step - loss: 0.0331 - accuracy: 0.9900 - lr: 3.8742e-04\nEpoch 11/40\n500/500 [==============================] - 127s 253ms/step - loss: 0.0344 - accuracy: 0.9896 - lr: 3.4868e-04\nEpoch 12/40\n500/500 [==============================] - 127s 253ms/step - loss: 0.0313 - accuracy: 0.9912 - lr: 3.1381e-04\nEpoch 13/40\n500/500 [==============================] - 129s 258ms/step - loss: 0.0320 - accuracy: 0.9911 - lr: 2.8243e-04\nEpoch 14/40\n500/500 [==============================] - 126s 253ms/step - loss: 0.0324 - accuracy: 0.9906 - lr: 2.5419e-04\nEpoch 15/40\n500/500 [==============================] - 139s 278ms/step - loss: 0.0295 - accuracy: 0.9929 - lr: 2.2877e-04\nEpoch 16/40\n500/500 [==============================] - 126s 252ms/step - loss: 0.0379 - accuracy: 0.9911 - lr: 2.0589e-04\nEpoch 17/40\n500/500 [==============================] - 127s 253ms/step - loss: 0.0282 - accuracy: 0.9914 - lr: 1.8530e-04\nEpoch 18/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0228 - accuracy: 0.9930 - lr: 1.6677e-04\nEpoch 19/40\n500/500 [==============================] - 126s 253ms/step - loss: 0.0294 - accuracy: 0.9912 - lr: 1.5009e-04\nEpoch 20/40\n500/500 [==============================] - 127s 253ms/step - loss: 0.0269 - accuracy: 0.9924 - lr: 1.3509e-04\nEpoch 21/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0307 - accuracy: 0.9915 - lr: 1.2158e-04\nEpoch 22/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0230 - accuracy: 0.9934 - lr: 1.0942e-04\nEpoch 23/40\n500/500 [==============================] - 127s 253ms/step - loss: 0.0195 - accuracy: 0.9942 - lr: 9.8477e-05\nEpoch 24/40\n500/500 [==============================] - 132s 264ms/step - loss: 0.0254 - accuracy: 0.9927 - lr: 8.8629e-05\nEpoch 25/40\n500/500 [==============================] - 127s 255ms/step - loss: 0.0215 - accuracy: 0.9933 - lr: 7.9766e-05\nEpoch 26/40\n500/500 [==============================] - 127s 253ms/step - loss: 0.0220 - accuracy: 0.9940 - lr: 7.1790e-05\nEpoch 27/40\n500/500 [==============================] - 128s 255ms/step - loss: 0.0215 - accuracy: 0.9940 - lr: 6.4611e-05\nEpoch 28/40\n500/500 [==============================] - 130s 260ms/step - loss: 0.0237 - accuracy: 0.9944 - lr: 5.8150e-05\nEpoch 29/40\n500/500 [==============================] - 139s 278ms/step - loss: 0.0224 - accuracy: 0.9933 - lr: 5.2335e-05\nEpoch 30/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0188 - accuracy: 0.9944 - lr: 4.7101e-05\nEpoch 31/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0251 - accuracy: 0.9946 - lr: 4.2391e-05\nEpoch 32/40\n500/500 [==============================] - 127s 255ms/step - loss: 0.0255 - accuracy: 0.9934 - lr: 3.8152e-05\nEpoch 33/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0206 - accuracy: 0.9950 - lr: 3.4337e-05\nEpoch 34/40\n500/500 [==============================] - 128s 255ms/step - loss: 0.0262 - accuracy: 0.9925 - lr: 3.0903e-05\nEpoch 35/40\n500/500 [==============================] - 128s 255ms/step - loss: 0.0228 - accuracy: 0.9926 - lr: 2.7813e-05\nEpoch 36/40\n500/500 [==============================] - 127s 255ms/step - loss: 0.0183 - accuracy: 0.9948 - lr: 2.5032e-05\nEpoch 37/40\n500/500 [==============================] - 128s 255ms/step - loss: 0.0172 - accuracy: 0.9958 - lr: 2.2528e-05\nEpoch 38/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0235 - accuracy: 0.9940 - lr: 2.0276e-05\nEpoch 39/40\n500/500 [==============================] - 128s 256ms/step - loss: 0.0219 - accuracy: 0.9944 - lr: 1.8248e-05\nEpoch 40/40\n500/500 [==============================] - 127s 254ms/step - loss: 0.0203 - accuracy: 0.9946 - lr: 1.6423e-05\n","name":"stdout"},{"output_type":"execute_result","execution_count":113,"data":{"text/plain":"<tensorflow.python.keras.callbacks.History at 0x7fa570bdd850>"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"result=model.predict(X_test)","execution_count":115,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"y_pred=pd.Series(np.argmax(result, axis=1), name='Label')\ny_pred","execution_count":116,"outputs":[{"output_type":"execute_result","execution_count":116,"data":{"text/plain":"0       0\n1       1\n2       2\n3       3\n4       4\n       ..\n9995    5\n9996    6\n9997    7\n9998    8\n9999    9\nName: Label, Length: 10000, dtype: int64"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.metrics import confusion_matrix, classification_report, accuracy_score\ncm = confusion_matrix(y_test, y_pred)\n","execution_count":117,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.figure(figsize=[7,6])\nsns.heatmap(cm, cmap=\"Reds\", annot=True, fmt='.0f')\nplt.show()","execution_count":118,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 504x432 with 2 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"acc=accuracy_score(y_test, y_pred)\nprint(acc)","execution_count":119,"outputs":[{"output_type":"stream","text":"0.9854\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"name":"python3","display_name":"Python 3","language":"python"},"language_info":{"name":"python","version":"3.7.6","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":4}