{"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_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# Introdução","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:49:27.746901Z","iopub.execute_input":"2022-03-11T02:49:27.747235Z","iopub.status.idle":"2022-03-11T02:49:27.751198Z","shell.execute_reply.started":"2022-03-11T02:49:27.747205Z","shell.execute_reply":"2022-03-11T02:49:27.750156Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Resumo:**\n\nO exemplo aqui desenvolvido tem como objetivo apresentar conceitos iniciais de implementação de redes neurais com python e tensorflow/keras. Esse modelo apresenta um modelo de Rede Neural Convolucional (CNN) básico que pode ser expandido mudando o número de neurônios e camadas. Em adaptações mais avançadas, pode-se estudar possibilidade de otimização de hyperparâmetros e outras técnincas como aumento de dados.\n\n**Não é objetivo nosso desenvolver e otimizar o modelo de classificação**. O exemplo tem objetivo meramente didático.\n\n---\n\n**Para saber mais:**\n* [Palestras e cursos do Ocean](http://www.oceanbrasil.com/)\n* François Chollet. Deep Learning with Python. Manning Publications, 2017.\n* Ian Goodfellow and Yoshua Bengio and Aaron Courville. [Deep Learning](https://www.deeplearningbook.org/). MIT Press, 2016.","metadata":{}},{"cell_type":"markdown","source":"# Bibliotecas e Dados","metadata":{}},{"cell_type":"code","source":"# Bibliotecas necessárias\n# Manipulação de dados\nimport pandas as pd\n# Redes Neurais\nfrom tensorflow import keras\nfrom tensorflow.keras.datasets import mnist\nfrom tensorflow.keras.models import Sequential\nfrom tensorflow.keras.layers import Dense, Dropout, Conv2D, MaxPooling2D, Flatten\nfrom tensorflow.keras.optimizers import RMSprop\n# Plot\nimport matplotlib.pyplot as plt\n%matplotlib inline\n# Avaliação\nfrom sklearn.metrics import classification_report, confusion_matrix\n","metadata":{"ExecuteTime":{"end_time":"2019-01-16T20:39:41.500258Z","start_time":"2019-01-16T20:39:39.261856Z"},"execution":{"iopub.status.busy":"2022-03-11T02:49:27.756983Z","iopub.execute_input":"2022-03-11T02:49:27.757841Z","iopub.status.idle":"2022-03-11T02:49:27.769677Z","shell.execute_reply.started":"2022-03-11T02:49:27.757726Z","shell.execute_reply":"2022-03-11T02:49:27.768849Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Lendo o dataset Kaggle\ntrain = pd.read_csv(\"../input/digit-recognizer/train.csv\")\n\n#Alternativa ler do próprio keras\n#(x_train, y_train), (x_test, y_test) = mnist.load_data()\n","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:49:27.771748Z","iopub.execute_input":"2022-03-11T02:49:27.772221Z","iopub.status.idle":"2022-03-11T02:49:29.875199Z","shell.execute_reply.started":"2022-03-11T02:49:27.772185Z","shell.execute_reply":"2022-03-11T02:49:29.874346Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Analisando o dataset\nprint(\"Quantidade de elementos de treino: {}\". format(len(train)))\nprint(train.head())","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:49:29.876671Z","iopub.execute_input":"2022-03-11T02:49:29.877159Z","iopub.status.idle":"2022-03-11T02:49:29.891664Z","shell.execute_reply.started":"2022-03-11T02:49:29.87712Z","shell.execute_reply":"2022-03-11T02:49:29.889503Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Separando x_train e y_train\nY = train[\"label\"]\nX = train.drop(labels = [\"label\"],axis = 1)\nprint(X.head())\n# Em formato numpy array de imagens 28 x 28\n#x = X.values.reshape(-1,28,28,1)\n#print(x[0])","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:49:29.89477Z","iopub.execute_input":"2022-03-11T02:49:29.895171Z","iopub.status.idle":"2022-03-11T02:49:29.988228Z","shell.execute_reply.started":"2022-03-11T02:49:29.895112Z","shell.execute_reply":"2022-03-11T02:49:29.9869Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(X.shape)","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:49:29.991266Z","iopub.execute_input":"2022-03-11T02:49:29.991659Z","iopub.status.idle":"2022-03-11T02:49:29.99698Z","shell.execute_reply.started":"2022-03-11T02:49:29.991605Z","shell.execute_reply":"2022-03-11T02:49:29.995095Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Numtendi nada!\n# Bora ver com matplotlib\nplt.imshow(X.values[0].reshape(28,28), cmap=plt.cm.binary)\nplt.show()\nprint('Label: {}'.format(Y[0]))","metadata":{"ExecuteTime":{"end_time":"2019-01-08T20:49:47.144492Z","start_time":"2019-01-08T20:49:47.016025Z"},"execution":{"iopub.status.busy":"2022-03-11T02:49:29.998182Z","iopub.execute_input":"2022-03-11T02:49:29.99844Z","iopub.status.idle":"2022-03-11T02:49:30.130144Z","shell.execute_reply.started":"2022-03-11T02:49:29.998415Z","shell.execute_reply":"2022-03-11T02:49:30.129174Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Transformando a imagem 2d em um numpy array (imagem 28*28)\nx = X.values.reshape(42000, 28, 28, 1)\n\n#Normalizando para valores entre 0 e 1\nx = x.astype('float32')\nx /= 255\n\n#print(x[0])","metadata":{"ExecuteTime":{"end_time":"2019-01-08T20:49:57.425537Z","start_time":"2019-01-08T20:49:57.272377Z"},"execution":{"iopub.status.busy":"2022-03-11T02:49:30.131502Z","iopub.execute_input":"2022-03-11T02:49:30.132046Z","iopub.status.idle":"2022-03-11T02:49:30.206637Z","shell.execute_reply.started":"2022-03-11T02:49:30.132006Z","shell.execute_reply":"2022-03-11T02:49:30.205817Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Vamos ajustar o formato da saida\nnum_classes = 10\n\n# Convertendo para um vetor de saida com 10 dimensoes\n# ex. 8 => [0,0,0,0,0,0,0,0,1,0]\ny = keras.utils.to_categorical(Y, num_classes)\nprint(y[0])","metadata":{"ExecuteTime":{"end_time":"2019-01-08T20:50:48.007428Z","start_time":"2019-01-08T20:50:47.986527Z"},"execution":{"iopub.status.busy":"2022-03-11T02:49:30.210199Z","iopub.execute_input":"2022-03-11T02:49:30.210467Z","iopub.status.idle":"2022-03-11T02:49:30.217731Z","shell.execute_reply.started":"2022-03-11T02:49:30.210441Z","shell.execute_reply":"2022-03-11T02:49:30.216723Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Separando uma parte para treino (90%) e outra para validação (10%)\nfrom sklearn.model_selection import train_test_split\nx_train, x_val, y_train, y_val = train_test_split(x, y, test_size = 0.1, random_state=5)\nprint('Qtde de treino: {}'.format(len(x_train)))\nprint('Qtde de validação: {}'.format(len(x_val)))","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:49:30.21915Z","iopub.execute_input":"2022-03-11T02:49:30.219716Z","iopub.status.idle":"2022-03-11T02:49:30.531131Z","shell.execute_reply.started":"2022-03-11T02:49:30.219676Z","shell.execute_reply":"2022-03-11T02:49:30.530012Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Criando e treinando o Modelo","metadata":{}},{"cell_type":"code","source":"# Criando o modelo Sequential\n# Sequential: Modelo Keras de ir adicionando camadas (como um lego)\n# Conv2D: Camada com kernels (filtros) que percorrem a imagem extraindo caracterísitcas (mapas de caracteísticas)\n# MaxPooling2D: Camada que reduz a dimensionalidade dos mapas de características 2D\n# Flatten: Camada que transforma um mapa de características 2D num vetor para classficador final\n# Dense: Camada onde todas as entradas estão conectadas em cada neurônio (totalmente conectada)\n# Dropout: Camada usa durante treino que descarta aleatoriamente um percentual de conexões (reduz overfitting)\n\n#Comentários Maria Gabriela Lima Damasceno Atividade 1 - CNN Simples com Keras\n#Foram aumentadas as camadas de Conv2D para 40 e 80, respectivamente.\n#Foram adcionadas 5 camadas Dense, onde as entradas são conectadas em cada neurônio, cada uma com 100, 120, 140, 160 e 180 neurônios, respectivamente.\n#A camada de Dropout foi reduzida para 0.1 e deixada após última camada Dense com ativação Relu para que fosse reduzido o overfitting.\n#Por fim, foi reduzido o parâmetro epochs para que o modelo pudesse convergir mais rapidamente para uma melhor performance. \n\nmodel = Sequential()\nmodel.add(Conv2D(40, kernel_size=(3, 3),\n                 activation='relu',\n                 input_shape=(28,28,1)))\nmodel.add(MaxPooling2D(pool_size=(2, 2)))\nmodel.add(Conv2D(80, kernel_size=(3,3), activation='relu'))\nmodel.add(Flatten())\nmodel.add(Dense(100, activation='relu'))\nmodel.add(Dense(120, activation='relu'))\nmodel.add(Dense(140, activation='relu'))\nmodel.add(Dense(160, activation='relu'))\nmodel.add(Dense(180, activation='relu'))\nmodel.add(Dropout(0.1))\nmodel.add(Dense(num_classes, activation='softmax'))\nmodel.summary()","metadata":{"ExecuteTime":{"end_time":"2019-01-08T21:38:00.233865Z","start_time":"2019-01-08T21:38:00.15079Z"},"scrolled":true,"execution":{"iopub.status.busy":"2022-03-11T02:49:30.534234Z","iopub.execute_input":"2022-03-11T02:49:30.534495Z","iopub.status.idle":"2022-03-11T02:49:30.638728Z","shell.execute_reply.started":"2022-03-11T02:49:30.534469Z","shell.execute_reply":"2022-03-11T02:49:30.637997Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Compila o modelo\nmodel.compile(loss='categorical_crossentropy',\n              optimizer=RMSprop(),\n              metrics=['accuracy'])","metadata":{"ExecuteTime":{"end_time":"2019-01-08T21:33:17.184644Z","start_time":"2019-01-08T21:33:17.129141Z"},"execution":{"iopub.status.busy":"2022-03-11T02:49:30.642028Z","iopub.execute_input":"2022-03-11T02:49:30.642286Z","iopub.status.idle":"2022-03-11T02:49:30.673429Z","shell.execute_reply.started":"2022-03-11T02:49:30.64226Z","shell.execute_reply":"2022-03-11T02:49:30.672687Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Treina com os parte dos dados\nbatch_size = 32\nepochs = 13\n\n#Salvar o melhor modelo\ncallbacks_list = [\n    keras.callbacks.ModelCheckpoint(\n        filepath='model.h5',\n        monitor='val_loss', save_best_only=True, verbose=1),\n    keras.callbacks.EarlyStopping(monitor='val_loss', patience=10,verbose=1)\n]\n\nhistory = model.fit(x_train, y_train,\n                    batch_size=batch_size,\n                    epochs=epochs,\n                    callbacks = callbacks_list,\n                    verbose=1,\n                    validation_data=(x_val, y_val))","metadata":{"ExecuteTime":{"end_time":"2019-01-08T21:36:15.501168Z","start_time":"2019-01-08T21:33:19.322895Z"},"execution":{"iopub.status.busy":"2022-03-11T02:49:30.676289Z","iopub.execute_input":"2022-03-11T02:49:30.67659Z","iopub.status.idle":"2022-03-11T02:50:43.330353Z","shell.execute_reply.started":"2022-03-11T02:49:30.676562Z","shell.execute_reply":"2022-03-11T02:50:43.329509Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#Vamos ver como foi o treino?\n\nfig, ax = plt.subplots(1,2, figsize=(16,8))\nax[0].plot(history.history['loss'], color='b', label=\"Training loss\")\nax[0].plot(history.history['val_loss'], color='r', label=\"validation loss\",axes =ax[0])\nlegend = ax[0].legend(loc='best', shadow=True)\n\nax[1].plot(history.history['accuracy'], color='b', label=\"Training accuracy\")\nax[1].plot(history.history['val_accuracy'], color='r',label=\"Validation accuracy\")\nlegend = ax[1].legend(loc='best', shadow=True)","metadata":{"ExecuteTime":{"end_time":"2019-01-08T21:36:20.243921Z","start_time":"2019-01-08T21:36:20.064201Z"},"execution":{"iopub.status.busy":"2022-03-11T02:50:43.333725Z","iopub.execute_input":"2022-03-11T02:50:43.333978Z","iopub.status.idle":"2022-03-11T02:50:43.62511Z","shell.execute_reply.started":"2022-03-11T02:50:43.333953Z","shell.execute_reply":"2022-03-11T02:50:43.624252Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from tensorflow.keras.models import load_model\n# Load the best saved model\nmodel = load_model('model.h5')","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:50:43.626608Z","iopub.execute_input":"2022-03-11T02:50:43.627306Z","iopub.status.idle":"2022-03-11T02:50:44.588485Z","shell.execute_reply.started":"2022-03-11T02:50:43.627262Z","shell.execute_reply":"2022-03-11T02:50:44.587703Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Testa\nscore = model.evaluate(x_val, y_val, verbose=0)\nprint('Test loss:', score[0])\nprint('Test accuracy:', score[1])","metadata":{"ExecuteTime":{"end_time":"2019-01-08T21:36:39.779842Z","start_time":"2019-01-08T21:36:38.946173Z"},"execution":{"iopub.status.busy":"2022-03-11T02:50:44.589882Z","iopub.execute_input":"2022-03-11T02:50:44.590229Z","iopub.status.idle":"2022-03-11T02:50:45.022222Z","shell.execute_reply.started":"2022-03-11T02:50:44.590192Z","shell.execute_reply":"2022-03-11T02:50:45.021384Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Testando uma entrada qualquer\nprint(y_train[10])\nprint(model.predict(x_train[10].reshape((1,28,28,1))))\nprint(model.predict_classes(x_train[10].reshape((1,28,28,1))))","metadata":{"ExecuteTime":{"end_time":"2019-01-08T21:06:04.363715Z","start_time":"2019-01-08T21:06:04.27849Z"},"execution":{"iopub.status.busy":"2022-03-11T02:50:45.024638Z","iopub.execute_input":"2022-03-11T02:50:45.02535Z","iopub.status.idle":"2022-03-11T02:50:45.145459Z","shell.execute_reply.started":"2022-03-11T02:50:45.025302Z","shell.execute_reply":"2022-03-11T02:50:45.144332Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Avaliando o Modelo","metadata":{}},{"cell_type":"code","source":"import itertools\n\n#Plot the confusion matrix. Set Normalize = True/False\ndef plot_confusion_matrix(cm, classes, normalize=True, title='Confusion matrix', cmap=plt.cm.Blues):\n    \"\"\"\n    This function prints and plots the confusion matrix.\n    Normalization can be applied by setting `normalize=True`.\n    \"\"\"\n    plt.figure(figsize=(10,10))\n    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n    plt.title(title)\n    plt.colorbar()\n    tick_marks = np.arange(len(classes))\n    plt.xticks(tick_marks, classes, rotation=45)\n    plt.yticks(tick_marks, classes)\n    if normalize:\n        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n        cm = np.around(cm, decimals=2)\n        cm[np.isnan(cm)] = 0.0\n    thresh = cm.max() / 2.\n    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n        plt.text(j, i, cm[i, j],\n                 horizontalalignment=\"center\",\n                 color=\"white\" if cm[i, j] > thresh else \"black\")\n    plt.tight_layout()\n    plt.ylabel('True label')\n    plt.xlabel('Predicted label')","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:50:45.147222Z","iopub.execute_input":"2022-03-11T02:50:45.147928Z","iopub.status.idle":"2022-03-11T02:50:45.165393Z","shell.execute_reply.started":"2022-03-11T02:50:45.147877Z","shell.execute_reply":"2022-03-11T02:50:45.164492Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Vendo alguns reports# Vendo alguns reports\n# Usando sklearn\nimport numpy as np\n\n# Classificando toda base de teste\ny_pred = model.predict_classes(x_val)\n# voltando pro formato de classes\ny_test_c = np.argmax(y_val, axis=1)\ntarget_names = ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']\n\n#Confution Matrix\ncm = confusion_matrix(y_test_c, y_pred)\nplot_confusion_matrix(cm, target_names, normalize=False, title='Confusion Matrix')\n\nprint('Classification Report')\nprint(classification_report(y_test_c, y_pred, target_names=target_names))","metadata":{"ExecuteTime":{"end_time":"2019-01-08T21:24:57.7654Z","start_time":"2019-01-08T21:24:57.220867Z"},"execution":{"iopub.status.busy":"2022-03-11T02:50:45.168282Z","iopub.execute_input":"2022-03-11T02:50:45.168576Z","iopub.status.idle":"2022-03-11T02:50:45.873588Z","shell.execute_reply.started":"2022-03-11T02:50:45.168548Z","shell.execute_reply":"2022-03-11T02:50:45.872719Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Gerando Saída","metadata":{}},{"cell_type":"code","source":"# Gerando saída para dataset de teste\n\n#Carrega dataset de teste\ntest = pd.read_csv(\"../input/digit-recognizer/test.csv\")\nprint(\"Qtde de testes: {}\".format(len(test)))\n# Bota no formato numpy e normaliza\nx_test = test.values.reshape(len(test),28,28,1)\nx_test = x_test.astype('float32')\nx_test /= 255\n\n# Faz classificação para dataset de teste\ny_pred = model.predict_classes(x_test)\n\n# Verficando algum exemplo\ni = 0\nplt.imshow(test.values[i].reshape(28,28), cmap=plt.cm.binary)\nplt.show()\nprint('Previsto: {}'.format(y_pred[i]))\n\n# Botando no formato de saída (competição Kaggle)\nresults = pd.Series(y_pred,name=\"Label\")\nsubmission = pd.concat([pd.Series(range(1,len(y_pred)+1),name = \"ImageId\"),results],axis = 1)\nprint(submission.head(10))\n#Salvando Arquivo\nsubmission.to_csv(\"mlp_mnist_v1.csv\",index=False)","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:50:45.877227Z","iopub.execute_input":"2022-03-11T02:50:45.877494Z","iopub.status.idle":"2022-03-11T02:50:48.881242Z","shell.execute_reply.started":"2022-03-11T02:50:45.877466Z","shell.execute_reply":"2022-03-11T02:50:48.880255Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Teste Adicional: Com ruído","metadata":{}},{"cell_type":"code","source":"#introduzindo ruido\nimport numpy as np\nmean = 0.\nstddev = 0.2\nnoise = np.random.normal(mean, stddev, (4200, 28, 28,1))\nx_te = x_val + noise\nx_te = np.clip(x_te, 0., 1.)\n\nplt.imshow(x_te.reshape(4200, 28,28)[0], cmap=plt.cm.binary)\nplt.show()\n\n# Testa\nscore = model.evaluate(x_te, y_val, verbose=0)\nprint('Test loss:', score[0])\nprint('Test accuracy:', score[1])","metadata":{"execution":{"iopub.status.busy":"2022-03-11T02:50:48.882716Z","iopub.execute_input":"2022-03-11T02:50:48.883026Z","iopub.status.idle":"2022-03-11T02:50:49.570601Z","shell.execute_reply.started":"2022-03-11T02:50:48.882998Z","shell.execute_reply":"2022-03-11T02:50:49.569737Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Com ruído, percebemos que a acurácia dos exemplos de validação caiu, mas muito menos que o modelo MLP clássico (ver exemplo de MLP).\nUm modelo Convolucional (CNN) captura melhor regiões, ou padrões espaciais de pixels, tornando o modelo mais robusto.","metadata":{}},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]}]}