{"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\nfrom sklearn.metrics import accuracy_score","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"datos=pd.read_csv('../input/digit-recognizer/train.csv')\ntest_data=pd.read_csv('../input/digit-recognizer/test.csv')\n\ninput_data=datos.iloc[:,1:]\noutput_data=datos.iloc[:,0]\n\nfrom sklearn.model_selection import train_test_split\n\ninput_data2=input_data+10*np.random.randn(42000,784)**2\ninput_data2=pd.concat([input_data2,np.round(2*(input_data2/255))/2])\noutput_data=pd.concat([output_data,output_data])\nX_train, X_test, y_train, y_test = train_test_split(input_data2/255, output_data, test_size=0.01, random_state=42)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def show_image(n):\n    img=input_data2.loc[n][:].iloc[1].values.reshape(28,28)\n\n    plt.title('El numero es: '+str(datos.loc[n][0]))\n    plt.imshow(img,cmap='binary')\n\nshow_image(14)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Tree desicion\n"},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn import tree\nclf = tree.DecisionTreeClassifier()\nclf = clf.fit(X_train, y_train)\n\ny_predict=clf.predict(X_test)\n\nfrom sklearn.metrics import confusion_matrix\nplt.matshow(confusion_matrix(y_test, y_predict))\naccuracy_score(y_test, y_predict)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.ensemble import RandomForestClassifier\nclf2 = RandomForestClassifier(max_depth=18, random_state=0)\nclf2.fit(X_train, y_train)\n\ny_predict2=clf2.predict(X_test)\n\nplt.matshow(confusion_matrix(y_test, y_predict2))\naccuracy_score(y_test, y_predict2)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from xgboost import XGBRegressor\nmy_model_1 = XGBRegressor(n_estimators=10, learning_rate=0.15,random_state=0)\nmy_model_1.fit(X_train, y_train)\n\ny_predict3=my_model_1.predict(X_test)\n\nplt.matshow(confusion_matrix(y_test, np.round(y_predict3)))\naccuracy_score(y_test, np.round(y_predict3))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"import tensorflow as tf\nfrom tensorflow import keras\n\nmodel = keras.Sequential([\n    keras.layers.Flatten(input_shape=(1, 784)),\n    keras.layers.Dense(80, activation='relu'),\n    keras.layers.Dropout(0.1),\n    keras.layers.Dense(10, activation='softmax')\n])\n\nmodel.compile(optimizer='adam',\n              loss='sparse_categorical_crossentropy',\n              metrics=['accuracy'])\n\nmodel.fit(X_train, y_train, epochs=10)\n\ny_predict4 = model.predict(X_test)\n\nplt.matshow(confusion_matrix(y_test, np.argmax(y_predict4,axis=1)))\naccuracy_score(y_test, np.argmax(y_predict4,axis=1))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model2 = keras.Sequential([\n#     keras.layers.Flatten(),\n    keras.Input(shape=(28, 28,1)),\n#     keras.layers.experimental.preprocessing.RandomRotation(0.05),\n    keras.layers.Conv2D(25, (3, 3), padding=\"valid\", activation=\"relu\"),\n    keras.layers.Conv2D(25, (10, 10), padding=\"valid\", activation=\"relu\"),\n    keras.layers.MaxPooling2D(),\n    keras.layers.Conv2D(25, (3, 3), padding=\"valid\", activation=\"relu\"),\n    keras.layers.MaxPooling2D(),\n    keras.layers.Conv2D(25, (1, 1), padding=\"valid\", activation=\"tanh\"),\n    keras.layers.Flatten(),\n    keras.layers.Dropout(0.6),\n    keras.layers.Dense(10, activation='softmax')\n])\n\nmodel2.summary()\nmodel2.compile(optimizer='adam',\n              loss='sparse_categorical_crossentropy',\n              metrics=['accuracy'])\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\nhistory=model2.fit(X_train.values.reshape((-1,28,28,1)), y_train.values.reshape(-1,1), epochs=100, batch_size=1000, validation_split=0.3)\n\ny_predict4 = model2.predict(X_test.iloc[:,:].values.reshape((-1,28,28,1)))\n\nplt.matshow(confusion_matrix(y_test, np.argmax(y_predict4,axis=1)))\naccuracy_score(y_test, np.argmax(y_predict4,axis=1))\n\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.plot(history.history['accuracy'])\nplt.plot(history.history['val_accuracy'])\nplt.legend(['acc','val'])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"faileture=np.argmax(y_predict4,axis=1)!=y_test\nprint(sum(faileture))\nfor i in range(sum(faileture)):\n    plt.figure()\n    plt.imshow(X_test[faileture].values.reshape((-1,28,28))[i],cmap='binary')\n    plt.title(str(np.argmax(y_predict4,axis=1)[i])+' vs '+str(y_test[faileture].iloc[i]))\n    plt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model2.fit(X_test[faileture].values.reshape((-1,28,28,1)), y_test[faileture], epochs=10)\n\ny_predict4 = model2.predict(X_test.iloc[:,:].values.reshape((-1,28,28,1)))\n\nplt.matshow(confusion_matrix(y_test, np.argmax(y_predict4,axis=1)))\naccuracy_score(y_test, np.argmax(y_predict4,axis=1))","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":4}