{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"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%matplotlib inline\nimport seaborn as sns\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 20GB 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","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2023-09-08T09:06:48.471303Z","iopub.execute_input":"2023-09-08T09:06:48.471774Z","iopub.status.idle":"2023-09-08T09:06:48.493090Z","shell.execute_reply.started":"2023-09-08T09:06:48.471737Z","shell.execute_reply":"2023-09-08T09:06:48.491823Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train = pd.read_csv(\"/kaggle/input/digit-recognizer/train.csv\")\ntest = pd.read_csv(\"/kaggle/input/digit-recognizer/test.csv\")","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:06:48.495844Z","iopub.execute_input":"2023-09-08T09:06:48.496647Z","iopub.status.idle":"2023-09-08T09:06:55.587242Z","shell.execute_reply.started":"2023-09-08T09:06:48.496613Z","shell.execute_reply":"2023-09-08T09:06:55.586223Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train.head()","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:06:55.588696Z","iopub.execute_input":"2023-09-08T09:06:55.589032Z","iopub.status.idle":"2023-09-08T09:06:55.615752Z","shell.execute_reply.started":"2023-09-08T09:06:55.588999Z","shell.execute_reply":"2023-09-08T09:06:55.614467Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(train.shape)\nprint('_'*60)\nprint(test.shape)","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:06:55.618992Z","iopub.execute_input":"2023-09-08T09:06:55.619353Z","iopub.status.idle":"2023-09-08T09:06:55.626482Z","shell.execute_reply.started":"2023-09-08T09:06:55.619321Z","shell.execute_reply":"2023-09-08T09:06:55.625386Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"img = np.reshape(train.iloc[0][1:].values.astype('uint8'), (28,28))\nplt.imshow(img)","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:06:55.628805Z","iopub.execute_input":"2023-09-08T09:06:55.629207Z","iopub.status.idle":"2023-09-08T09:06:55.918488Z","shell.execute_reply.started":"2023-09-08T09:06:55.629183Z","shell.execute_reply":"2023-09-08T09:06:55.917405Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from keras.datasets import mnist\n\n(X_train1, y_train1), (X_test1, y_test1) = mnist.load_data()\n\nX_train1 = np.concatenate([X_train1, X_test1], axis=0)\ny_train1 = np.concatenate([y_train1, y_test1], axis=0)\n\nX_train1 = X_train1.reshape(-1, 28*28)\n\nprint('_'*60)\nprint(X_train1.shape)\nprint('_'*60)\nprint(y_train1.shape)","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:06:55.919962Z","iopub.execute_input":"2023-09-08T09:06:55.920308Z","iopub.status.idle":"2023-09-08T09:07:04.671118Z","shell.execute_reply.started":"2023-09-08T09:06:55.920276Z","shell.execute_reply":"2023-09-08T09:07:04.669893Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"normalization","metadata":{}},{"cell_type":"code","source":"X_train = train.drop([\"label\"], axis = 1)\ny_train = train[\"label\"]\n\nX_train = X_train / 255.0\nX_train1 = X_train1 / 255.0\n\nX_train = np.concatenate([X_train.values, X_train1])\ny_train = np.concatenate([y_train, y_train1])\n\nX_train = X_train.reshape(-1,28,28,1)\n\ntest = test / 255.0\ntest = test.values.reshape(-1,28,28,1)\n\nprint('_'*60)\nprint(X_train.shape)\nprint('_'*60)\nprint(y_train.shape)\nprint('_'*60)\nprint(test.shape)","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:07:04.672816Z","iopub.execute_input":"2023-09-08T09:07:04.673696Z","iopub.status.idle":"2023-09-08T09:07:05.348306Z","shell.execute_reply.started":"2023-09-08T09:07:04.673657Z","shell.execute_reply":"2023-09-08T09:07:05.347336Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"one hot encoding","metadata":{}},{"cell_type":"code","source":"from keras.utils.np_utils import to_categorical\n\ny_train = to_categorical(y_train, num_classes = 10)\ny_train","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:07:05.349981Z","iopub.execute_input":"2023-09-08T09:07:05.350324Z","iopub.status.idle":"2023-09-08T09:07:05.360131Z","shell.execute_reply.started":"2023-09-08T09:07:05.350291Z","shell.execute_reply":"2023-09-08T09:07:05.359043Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from sklearn.model_selection import train_test_split\n\nX_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size = 0.1, random_state=2)\n\nprint('_'*60)\nprint(X_train.shape)\nprint('_'*60)\nprint(X_val.shape)\nprint('_'*60)\nprint(y_train.shape)\nprint('_'*60)\nprint(y_val.shape)","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:07:05.362011Z","iopub.execute_input":"2023-09-08T09:07:05.362908Z","iopub.status.idle":"2023-09-08T09:07:05.728972Z","shell.execute_reply.started":"2023-09-08T09:07:05.362872Z","shell.execute_reply":"2023-09-08T09:07:05.727844Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"CNN\n\n\ninput 28x28x1\n\ncon2d + batch \\\ncon2d + batch \\\nmaxpool2d + dropout\n\ncon2d + batch \\\ncon2d + batch \\\nmaxpool2d + dropout\n\ncon2d + batch + dropout\n\nflatten + Dense x 256 + batch + dropout\n\nDense x 10","metadata":{}},{"cell_type":"code","source":"from keras.models import Sequential\nfrom keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPool2D, BatchNormalization\n\nmodel = Sequential()\n\nmodel.add(Conv2D(filters=64, kernel_size=(5,5) ,padding='Same', activation='relu', input_shape=(28,28,1)))\nmodel.add(BatchNormalization())\n\nmodel.add(Conv2D(filters=64, kernel_size=(5,5), padding='Same', activation='relu'))\nmodel.add(BatchNormalization())\n\nmodel.add(MaxPool2D(pool_size=(2,2)))\nmodel.add(Dropout(0.25))\n\nmodel.add(Conv2D(filters=64, kernel_size=(3,3),padding='Same', activation='relu'))\nmodel.add(BatchNormalization())\n\nmodel.add(Conv2D(filters=64, kernel_size=(3,3),padding='Same', activation='relu'))\nmodel.add(BatchNormalization())\n\nmodel.add(MaxPool2D(pool_size=(2,2), strides=(2,2)))\nmodel.add(Dropout(0.25))\n\nmodel.add(Conv2D(filters=64, kernel_size=(3,3), padding='Same',  activation='relu'))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.25))\n\nmodel.add(Flatten())\nmodel.add(Dense(256, activation=\"relu\"))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.25))\n\nmodel.add(Dense(10, activation = \"softmax\"))","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:07:05.732791Z","iopub.execute_input":"2023-09-08T09:07:05.733166Z","iopub.status.idle":"2023-09-08T09:07:08.413538Z","shell.execute_reply.started":"2023-09-08T09:07:05.733127Z","shell.execute_reply":"2023-09-08T09:07:08.412555Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from keras.optimizers import RMSprop\n\noptimizer = RMSprop(lr=0.001, rho=0.9, epsilon=1e-08, decay=0.0)\nmodel.compile(optimizer=optimizer , loss=\"categorical_crossentropy\", metrics=[\"accuracy\"])","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:07:08.415004Z","iopub.execute_input":"2023-09-08T09:07:08.415338Z","iopub.status.idle":"2023-09-08T09:07:08.436075Z","shell.execute_reply.started":"2023-09-08T09:07:08.415306Z","shell.execute_reply":"2023-09-08T09:07:08.434328Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from keras.callbacks import ReduceLROnPlateau\n\nepochs = 50\nbatch_size = 128\nlearning_rate_reduction = ReduceLROnPlateau(monitor='val_acc', patience=3, \\\n                                            verbose=1, factor=0.5, min_lr=0.00001)\n\nhistory = model.fit(X_train, y_train, epochs=epochs,validation_data=(X_val, y_val), \\\n                    verbose=2, steps_per_epoch=X_train.shape[0]//batch_size, \\\n                    callbacks=[learning_rate_reduction], \\\n                    validation_steps=X_val.shape[0]//batch_size)","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:07:08.437630Z","iopub.execute_input":"2023-09-08T09:07:08.437968Z","iopub.status.idle":"2023-09-08T09:17:32.910902Z","shell.execute_reply.started":"2023-09-08T09:07:08.437936Z","shell.execute_reply":"2023-09-08T09:17:32.909888Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"prediction = model.predict(test)","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:17:32.912869Z","iopub.execute_input":"2023-09-08T09:17:32.913242Z","iopub.status.idle":"2023-09-08T09:17:35.952899Z","shell.execute_reply.started":"2023-09-08T09:17:32.913207Z","shell.execute_reply":"2023-09-08T09:17:35.951773Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"prediction = [i.argmax() for i in prediction]","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:17:35.954425Z","iopub.execute_input":"2023-09-08T09:17:35.954780Z","iopub.status.idle":"2023-09-08T09:17:35.987536Z","shell.execute_reply.started":"2023-09-08T09:17:35.954747Z","shell.execute_reply":"2023-09-08T09:17:35.986586Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission = pd.DataFrame({\"imageid\": range(1, len(prediction) + 1), \"Label\": prediction})\nsubmission.to_csv(\"submission.csv\", index=False)\nsubmission.head()","metadata":{"execution":{"iopub.status.busy":"2023-09-08T09:17:35.989169Z","iopub.execute_input":"2023-09-08T09:17:35.989535Z","iopub.status.idle":"2023-09-08T09:17:36.210551Z","shell.execute_reply.started":"2023-09-08T09:17:35.989502Z","shell.execute_reply":"2023-09-08T09:17:36.209539Z"},"trusted":true},"execution_count":null,"outputs":[]}],"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"}}