{"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 in \n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the \"../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# Any results you write to the current directory are saved as output.","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"import tensorflow as tf\nfrom tensorflow import keras\nfrom tensorflow.keras import layers\nimport numpy as np\nimport pandas as pd\nimport os\nimport cv2\nfrom sklearn.utils import shuffle\nimport matplotlib.pyplot as plt\nfrom sklearn.model_selection import train_test_split as tts\nimport seaborn as sns\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_train=pd.read_csv('/kaggle/input/Kannada-MNIST/train.csv')\ny_train=x_train[['label']]\ny_train.shape\nx_train.drop(labels='label',axis=1,inplace=True)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_train=np.array(x_train).reshape(len(x_train),28,28,1)\nx_train=x_train/255.0\nprint(\"shape of training data=\",x_train.shape)\ny_train=np.array(y_train)\nprint(\"shape of training labels=\",y_train.shape)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"x_train,x_test,y_train,y_val=tts(x_train,y_train,test_size=0.2,random_state=42)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(\"training data shape=\",x_train.shape)\nprint(\"testing data shape=\",x_test.shape)\nprint(\"training labels shape=\",y_train.shape)\nprint(\"testing labels shape=\",y_val.shape)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_data=tf.data.Dataset.from_tensor_slices((x_train,y_train)).shuffle(10000).batch(4096)\ntest_data=tf.data.Dataset.from_tensor_slices((x_test,y_val)).shuffle(10000).batch(4096)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model=keras.Sequential()\nmodel.add(layers.Conv2D(32,(3,3),activation='relu',input_shape=(28,28,1)))\nmodel.add(layers.Conv2D(64,(3,3),activation='relu',padding='same',kernel_regularizer='l2'))\n\nmodel.add(layers.MaxPooling2D((2,2)))\nmodel.add(layers.Conv2D(64,(3,3),activation='relu',padding='same',kernel_regularizer='l2'))\nmodel.add(layers.BatchNormalization())\n\n\nmodel.add(layers.Conv2D(64,(3,3),activation='relu',padding='same',kernel_regularizer='l2'))\nmodel.add(layers.MaxPooling2D((2,2)))\n\n\nmodel.add(layers.Flatten())\n\nmodel.add(layers.Dense(512,activation='relu'))\nmodel.add(layers.Dropout(0.2))\nmodel.add(layers.Dense(512,activation='relu'))\nmodel.add(layers.Dropout(0.2))\nmodel.add(layers.Dense(512,activation='relu'))\nmodel.add(layers.Dropout(0.5))\n\n\nmodel.add(layers.Dense(10,activation='softmax'))\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.summary()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from tensorflow import keras\nreduce_lr = keras.callbacks.ReduceLROnPlateau(monitor='val_loss', factor=0.2,patience=3, min_lr=0.001)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.compile(loss='sparse_categorical_crossentropy',optimizer='adam',metrics=['accuracy'])\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"tf.keras.backend.clear_session()\nmodel.fit(train_data,epochs=10,callbacks=[reduce_lr])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.compile(loss='sparse_categorical_crossentropy',optimizer=tf.keras.optimizers.Adam(lr=0.0001),metrics=['accuracy'])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.fit(train_data,epochs=20,initial_epoch=10,callbacks=[reduce_lr])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.compile(loss='sparse_categorical_crossentropy',optimizer=tf.keras.optimizers.Adam(lr=0.00001),metrics=['accuracy'])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.fit(train_data,epochs=40,initial_epoch=20,callbacks=[reduce_lr])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.compile(loss='sparse_categorical_crossentropy',optimizer=tf.keras.optimizers.Adam(lr=0.00001,amsgrad=True),metrics=['accuracy'])\nmodel.fit(train_data,initial_epoch=40,epochs=60,callbacks=[reduce_lr])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.compile(loss='sparse_categorical_crossentropy',optimizer=tf.keras.optimizers.Adam(lr=0.000005,amsgrad=True),metrics=['accuracy'])\nmodel.fit(train_data,initial_epoch=60,epochs=80)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.compile(loss='sparse_categorical_crossentropy',optimizer=tf.keras.optimizers.Adam(lr=0.0000001,amsgrad=True),metrics=['accuracy'])\nmodel.fit(train_data,initial_epoch=80,epochs=100)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.evaluate(test_data)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\nplt.plot(model.history.history['accuracy'])\n#plt.plot(model.history.history['val_accuracy'])\nplt.xlabel('epochs')\nplt.ylabel('accuracy')\nplt.legend(['training_acc'])\nplt.show()\nplt.savefig('accuracy_plots.jpg')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.history.history.keys()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.plot(model.history.history['loss'])\n#plt.plot(model.history.history['val_loss'])\nplt.xlabel('epochs')\nplt.ylabel('loss')\nplt.title('loss of data over epochs')\nplt.legend(['training_loss','testing_loss'])\nplt.show()\nplt.savefig('loss_plts.jpg')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submit_x=pd.read_csv('/kaggle/input/Kannada-MNIST/test.csv')\nids=submit_x[['id']]\nsubmit_x.drop(labels='id',axis=1,inplace=True)\nsubmit_x=np.array(submit_x).reshape(len(submit_x),28,28,1)\nsubmit_x=submit_x/255.0\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submit_y=model.predict(submit_x)\nY=[]\nfor row in submit_y:\n    Y.append(np.argmax(row))\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"Y=np.array(Y)\nY=pd.DataFrame(Y)\nY.rename(columns={0:'label'},inplace=True)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"final_file=pd.concat([ids,Y],axis=1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"final_file.to_csv('submission.csv',index=False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":1}