{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"import numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"train = pd.read_csv('/kaggle/input/digit-recognizer/train.csv') #Reading Data\nX_test = pd.read_csv('/kaggle/input/digit-recognizer/test.csv')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train.shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train.shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_val = train.iloc[30000:,1:]\ny_val = train.iloc[30000:,0:1]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_test.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_train = train.iloc[:30000, 1:] # Splitting X & y\ny_train = train.iloc[:30000, 0:1]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Encoding y_train for CNN\nfrom sklearn.preprocessing import OneHotEncoder\nonehotencoder = OneHotEncoder()\ny_train = onehotencoder.fit_transform(y_train).toarray()\ny_train","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"X_train.max().max() # Hence Maximum Pixel value is 255.","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Normalising Training Set\nX_train /= 255\n\nprint(X_train.max().max())\nprint(X_train.min().min())","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Reshaping X_train for CNN...\nX_train = np.array(X_train)\nX_train = X_train.reshape((X_train.shape[0],28,28,1))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Now we are ready to begin with are Convulutional Nueral Network....\nfrom tensorflow.keras.models import Sequential\nfrom tensorflow.keras.layers import Dense, Dropout, Conv2D, MaxPool2D, Flatten","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Initialising CNN\ncnn = Sequential()\n\n# Adding Conv2D layer and MaxPool2D Layer\ncnn.add(Conv2D(filters=32, kernel_size=3, activation='relu', input_shape=(28, 28, 1)))\ncnn.add(MaxPool2D(pool_size=(2,2), strides=(2,2)))\n\n# Adding another Conv2D layer and MaxPool2D Layer\ncnn.add(Conv2D(filters=32, kernel_size=3, activation='relu'))        \ncnn.add(MaxPool2D(pool_size=(2,2), strides=(2,2)))\n\n# Flattening\ncnn.add(Flatten())\n\n# Adding ANN Hidden Layer\ncnn.add(Dense(units=128, activation='relu'))\n\n# Adding Output Layer using Softmax Activation Function\ncnn.add(Dense(units = 10,activation='softmax'))\n\n# Compiling NN\ncnn.compile(loss='categorical_crossentropy',optimizer='adam',metrics=['accuracy'])\n\n# Fiiting our NN to train set\ncnn.fit(X_train, y_train, epochs = 25, batch_size = 512)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Making Test Set ready for prediction...\nX_val /= 255\nX_val = np.array(X_val)\nX_val = X_val.reshape((X_val.shape[0],28,28,1))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Randomly Picking Nos. and predicting them\nimport matplotlib.pyplot as plt\nrandom_selection=np.random.randint(0,X_val.shape[0],size=1)\nrandom_sample=X_val[random_selection]\nprint('Prediction:')\nprint(cnn.predict_classes(random_sample.reshape(1,28,28,1))[0])\nplt.imshow(random_sample.reshape(28,28),cmap='binary')\nplt.show","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#  And Finally predicting\ntest_pred = cnn.predict_classes(X_val)\ntest_pred","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.metrics import confusion_matrix, accuracy_score\ncm = confusion_matrix(np.array(y_val), test_pred)\nprint(\"---------- Confusion Matrix ----------\")\nprint(cm)\nprint(\"---------- Accuracy ----------\")\nprint(accuracy_score(np.array(y_val), test_pred)) # Calculating Accuracy Over Test set","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Making submission...\nimg_id = np.arange(1, test_pred.size+1)\noutput=pd.DataFrame({'ImageId':img_id,'Label':test_pred})\noutput.to_csv('my_submission.csv',index=False)","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":4}