{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.13","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"gpu","dataSources":[{"sourceId":3004,"databundleVersionId":861823,"sourceType":"competition"}],"dockerImageVersionId":30747,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import warnings\nwarnings.filterwarnings(\"ignore\")\n\n# Import the necessary libs\nimport pandas as pd\nimport numpy as np\nimport os\n\nimport matplotlib.pyplot as plt\nimport matplotlib.image as mpimg\nimport seaborn as sns\n%matplotlib inline\n\nnp.random.seed(2)\n\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import confusion_matrix\nimport itertools\n\nfrom keras.utils import to_categorical # convert to one-hot-encoding\nfrom keras.models import Sequential\nfrom keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPool2D, BatchNormalization, Input\nfrom keras.optimizers import RMSprop\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator\nfrom keras.callbacks import ReduceLROnPlateau, Callback\nfrom keras.datasets import mnist\nfrom keras.utils import plot_model\n\nfrom IPython.display import Image\nimport tensorflow as tf\n\nsns.set(style='white', context='notebook', palette='deep')","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2024-08-01T00:39:25.233368Z","iopub.execute_input":"2024-08-01T00:39:25.234061Z","iopub.status.idle":"2024-08-01T00:39:39.522725Z","shell.execute_reply.started":"2024-08-01T00:39:25.234033Z","shell.execute_reply":"2024-08-01T00:39:39.521703Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Data Loading & Preparation\n----","metadata":{}},{"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)\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":{"execution":{"iopub.status.busy":"2024-08-01T00:39:39.535217Z","iopub.execute_input":"2024-08-01T00:39:39.535562Z","iopub.status.idle":"2024-08-01T00:39:39.544376Z","shell.execute_reply.started":"2024-08-01T00:39:39.535528Z","shell.execute_reply":"2024-08-01T00:39:39.543264Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train = pd.read_csv('../input/digit-recognizer/train.csv')\ntest = pd.read_csv('../input/digit-recognizer/test.csv')\nsample_submission = pd.read_csv('../input/digit-recognizer/sample_submission.csv')\n\nprint(\"Data are Ready!!\")\nprint(f\"Training data size is {train.shape}\\nTesting data size is {test.shape}\")","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:39.545362Z","iopub.execute_input":"2024-08-01T00:39:39.545871Z","iopub.status.idle":"2024-08-01T00:39:44.790205Z","shell.execute_reply.started":"2024-08-01T00:39:39.545839Z","shell.execute_reply":"2024-08-01T00:39:44.789266Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"Y_train = train[\"label\"]\nX_train = train.drop(labels = [\"label\"], axis = 1) ","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:44.791205Z","iopub.execute_input":"2024-08-01T00:39:44.791486Z","iopub.status.idle":"2024-08-01T00:39:44.878386Z","shell.execute_reply.started":"2024-08-01T00:39:44.791454Z","shell.execute_reply":"2024-08-01T00:39:44.877627Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Print data histogram to check if data targets/labels are balanced\nsns.countplot(x=Y_train)\nplt.xlabel('Classes')\nplt.ylabel('Frequency')\nplt.title('Class Distribution in Y_train')\nplt.show()\n","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:44.879460Z","iopub.execute_input":"2024-08-01T00:39:44.879801Z","iopub.status.idle":"2024-08-01T00:39:45.238953Z","shell.execute_reply.started":"2024-08-01T00:39:44.879777Z","shell.execute_reply":"2024-08-01T00:39:45.238050Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"Y_train_full = train[\"label\"]\nX_train_full = train.drop(labels = [\"label\"], axis = 1) \n\nX_train_full = X_train_full.values.reshape(-1,28,28,1)\ntest = test.values.reshape(-1,28,28,1)\n\nprint(\"Size of Dataset: \" , len(X_train_full))\n\nsplit_ratio = 0.1\nrandom_seed = 2\n\n# Split the train and the validation set for the fitting\nX_train, X_val, Y_train, Y_val = train_test_split(X_train_full, Y_train_full,\n                                                  test_size=split_ratio, random_state=random_seed)\nX_test = test","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:45.240046Z","iopub.execute_input":"2024-08-01T00:39:45.240323Z","iopub.status.idle":"2024-08-01T00:39:45.705043Z","shell.execute_reply.started":"2024-08-01T00:39:45.240300Z","shell.execute_reply":"2024-08-01T00:39:45.704148Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"random_idx = np.random.choice(X_train_full.shape[0], 15, replace=False)\nfor i, idx in enumerate(random_idx):\n    plt.subplot(3, 5, i+1)\n    plt.imshow(X_train_full[idx, :, :])\n    plt.xticks([])  # Remove coordinates from axes\n    plt.yticks([])\n    plt.title(Y_train_full[idx])","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:45.706139Z","iopub.execute_input":"2024-08-01T00:39:45.706393Z","iopub.status.idle":"2024-08-01T00:39:46.617064Z","shell.execute_reply.started":"2024-08-01T00:39:45.706370Z","shell.execute_reply":"2024-08-01T00:39:46.616097Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_train = np.array(X_train)\nX_val = np.array(X_val)\nX_test = np.array(X_test)\n\n# # Padding the images by 2 pixels since in the paper input images were 32x32\n# X_train = np.pad(X_train, ((0,0),(2,2),(2,2),(0,0)), 'constant')\n# X_val = np.pad(X_val, ((0,0),(2,2),(2,2),(0,0)), 'constant')\n# X_test = np.pad(X_test, ((0,0),(2,2),(2,2),(0,0)), 'constant')\n\n# # Standardization\n# mean_px = X_train.mean().astype(np.float32)\n# std_px = X_train.std().astype(np.float32)\n# X_train = (X_train - mean_px)/(std_px)\n\n# mean_px = X_val.mean().astype(np.float32)\n# std_px = X_val.std().astype(np.float32)\n# X_val = (X_val - mean_px)/(std_px)\n\n# mean_px = X_test.mean().astype(np.float32)\n# std_px = X_test.std().astype(np.float32)\n# X_test = (X_test - mean_px)/(std_px)\n\n# Normalization\nX_train = X_train.astype('float32') / 255.0\nX_val = X_val.astype('float32') / 255.0\nX_test = X_test.astype('float32') / 255.0\n\n# One-hot encoding the labels\nY_train = to_categorical(Y_train, num_classes = 10)\nY_val = to_categorical(Y_val, num_classes = 10)","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:46.621888Z","iopub.execute_input":"2024-08-01T00:39:46.622588Z","iopub.status.idle":"2024-08-01T00:39:46.896345Z","shell.execute_reply.started":"2024-08-01T00:39:46.622554Z","shell.execute_reply":"2024-08-01T00:39:46.895511Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Print out the type\nprint(\"Train Images Type:\", type(X_train))\nprint(\"Train Labels Type:\", type(Y_train))\nprint(\"Validation Images Type:\", type(X_val))\nprint(\"Validation Labels Type:\", type(Y_val))\nprint(\"Test Images Type:\",type(X_test))\n\n# Print out the shapes\nprint(\"\\nTrain Images Shape:\", X_train.shape)\nprint(\"Train Labels Shape:\", Y_train.shape)\nprint(\"Validation Images Shape:\", X_val.shape)\nprint(\"Validation Labels Shape:\", Y_val.shape)\nprint(\"Test Images Shape:\", X_test.shape)","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:46.897482Z","iopub.execute_input":"2024-08-01T00:39:46.897867Z","iopub.status.idle":"2024-08-01T00:39:46.905697Z","shell.execute_reply.started":"2024-08-01T00:39:46.897835Z","shell.execute_reply":"2024-08-01T00:39:46.904610Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model Creation\n----","metadata":{}},{"cell_type":"code","source":"input_shape = (28, 28, 1)","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:46.906836Z","iopub.execute_input":"2024-08-01T00:39:46.907130Z","iopub.status.idle":"2024-08-01T00:39:46.913676Z","shell.execute_reply.started":"2024-08-01T00:39:46.907106Z","shell.execute_reply":"2024-08-01T00:39:46.912827Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#Creating CNN model\n\"\"\"\n  [[Conv2D->relu]*2 -> BatchNormalization -> MaxPool2D -> Dropout]*2 -> \n  [Conv2D->relu]*2 -> BatchNormalization -> Dropout -> \n  Flatten -> Dense -> BatchNormalization -> Dropout -> Out\n\"\"\"\nmodel = Sequential()\n\nmodel.add(Input(shape=input_shape))\nmodel.add(Conv2D(filters = 64, kernel_size = (5,5),padding = 'Same', activation ='relu'))\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())\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":"2024-08-01T00:39:46.914857Z","iopub.execute_input":"2024-08-01T00:39:46.915128Z","iopub.status.idle":"2024-08-01T00:39:47.832980Z","shell.execute_reply.started":"2024-08-01T00:39:46.915106Z","shell.execute_reply":"2024-08-01T00:39:47.831990Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# visualize model\nplot_model(model, to_file='model.png', show_shapes=True, show_layer_names=True)\nImage(\"model.png\")","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:47.834144Z","iopub.execute_input":"2024-08-01T00:39:47.834406Z","iopub.status.idle":"2024-08-01T00:39:48.809636Z","shell.execute_reply.started":"2024-08-01T00:39:47.834383Z","shell.execute_reply":"2024-08-01T00:39:48.808702Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model.summary()","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:48.810929Z","iopub.execute_input":"2024-08-01T00:39:48.811216Z","iopub.status.idle":"2024-08-01T00:39:48.852375Z","shell.execute_reply.started":"2024-08-01T00:39:48.811192Z","shell.execute_reply":"2024-08-01T00:39:48.851546Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Define optimizer with correct argument\noptimizer = RMSprop(learning_rate=0.001, rho=0.9, epsilon=1e-08, decay=0.0)\n\n# Compile the model\nmodel.compile(optimizer=optimizer,\n              loss=\"categorical_crossentropy\",\n              metrics=[\"accuracy\"])\n\n# Define learning rate schedule\nvariable_lr = ReduceLROnPlateau(monitor='val_accuracy', \n                                patience=3, \n                                verbose=1, \n                                factor=0.5, \n                                min_lr=0.00001)\n\n# Model Settings\nepochs = 50\nbatch_size = 128\n","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:48.853588Z","iopub.execute_input":"2024-08-01T00:39:48.853949Z","iopub.status.idle":"2024-08-01T00:39:48.871167Z","shell.execute_reply.started":"2024-08-01T00:39:48.853917Z","shell.execute_reply":"2024-08-01T00:39:48.870167Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Data Augmentation\n----","metadata":{}},{"cell_type":"code","source":"# Data Augmentation \ndatagen = ImageDataGenerator(\n        featurewise_center=False, # set input mean to 0 over the dataset\n        samplewise_center=False,  # set each sample mean to 0\n        featurewise_std_normalization=False,  # divide inputs by std of the dataset\n        samplewise_std_normalization=False,  # divide each input by its std\n        zca_whitening=False,  # apply ZCA whitening\n        rotation_range=10,  # randomly rotate images in the range (degrees, 0 to 180)\n        zoom_range = 0.1, # Randomly zoom image \n        width_shift_range=0.1,  # randomly shift images horizontally (fraction of total width)\n        height_shift_range=0.1,  # randomly shift images vertically (fraction of total height)\n        horizontal_flip=False,  # randomly flip images\n        vertical_flip=False)  # randomly flip images","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:48.872206Z","iopub.execute_input":"2024-08-01T00:39:48.872442Z","iopub.status.idle":"2024-08-01T00:39:48.877482Z","shell.execute_reply.started":"2024-08-01T00:39:48.872421Z","shell.execute_reply":"2024-08-01T00:39:48.876693Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model Training\n----","metadata":{}},{"cell_type":"code","source":"# Initialize a list to store learning rates\nlearning_rates = []\n\n# Custom callback to log learning rates\nclass LogLearningRate(Callback):\n    def on_epoch_end(self, epoch, logs=None):\n        lr = self.model.optimizer.learning_rate.numpy()\n        learning_rates.append(lr)\n\n# Add the custom callback to the list of callbacks\nlog_lr = LogLearningRate()\n\n# Compute the internal data augmented stats related to the training dataset\n# datagen.fit(X_train)\n\n# Fit the model on the batches (images and labels) generated by datagen.flow() on the fly\ntrain_gen = datagen.flow(X_train,Y_train, batch_size=batch_size)\n\n# Train the model with the custom callback and augmented data\nhistory = model.fit(train_gen, \n                    epochs=epochs,\n                    validation_data=(X_val, Y_val),\n#                     verbose=2,\n                    steps_per_epoch=X_train.shape[0] // batch_size,\n                    callbacks=[variable_lr, log_lr],\n#                     validation_steps=X_val.shape[0] // batch_size\n                   )","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:39:48.878631Z","iopub.execute_input":"2024-08-01T00:39:48.878970Z","iopub.status.idle":"2024-08-01T00:44:57.410939Z","shell.execute_reply.started":"2024-08-01T00:39:48.878947Z","shell.execute_reply":"2024-08-01T00:44:57.410144Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model Evaluation\n----","metadata":{}},{"cell_type":"code","source":"# Plotting the training and validation loss and accuracy\ndef plot_history(history, learning_rates):\n    fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 4))\n\n    # Plotting loss\n    ax1.plot(history.history['loss'], label='Training Loss')\n    ax1.plot(history.history['val_loss'], label='Validation Loss')\n    ax1.set_title('Loss')\n    ax1.set_xlabel('Epoch')\n    ax1.set_ylabel('Loss')\n    ax1.legend()\n\n    # Plotting accuracy\n    ax2.plot(history.history['accuracy'], label='Training Accuracy')\n    ax2.plot(history.history['val_accuracy'], label='Validation Accuracy')\n    ax2.set_title('Accuracy')\n    ax2.set_xlabel('Epoch')\n    ax2.set_ylabel('Accuracy')\n    ax2.legend()\n\n    plt.show()\n\n    # Plotting learning rate\n    plt.figure(figsize=(6, 4))\n    plt.plot(learning_rates, label='Learning Rate')\n    plt.title('Learning Rate Schedule')\n    plt.xlabel('Epoch')\n    plt.ylabel('Learning Rate')\n    plt.legend()\n    plt.show()\n\n# Call the function to plot the history and learning rate\nplot_history(history, learning_rates)","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:44:57.412414Z","iopub.execute_input":"2024-08-01T00:44:57.413191Z","iopub.status.idle":"2024-08-01T00:44:58.269720Z","shell.execute_reply.started":"2024-08-01T00:44:57.413154Z","shell.execute_reply":"2024-08-01T00:44:58.268806Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model Predictions + Create Submission File\n----","metadata":{}},{"cell_type":"code","source":"sample_submission.head()","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:44:58.271101Z","iopub.execute_input":"2024-08-01T00:44:58.271549Z","iopub.status.idle":"2024-08-01T00:44:58.284315Z","shell.execute_reply.started":"2024-08-01T00:44:58.271509Z","shell.execute_reply":"2024-08-01T00:44:58.283440Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"predictions = model.predict(X_test)\npredicted_labels = np.argmax(predictions, axis=1)\nsubmission = pd.DataFrame({\n    'ImageId': range(1, len(predicted_labels) + 1), \n    'Label': predicted_labels\n})\nsubmission.head()","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:44:58.285806Z","iopub.execute_input":"2024-08-01T00:44:58.286127Z","iopub.status.idle":"2024-08-01T00:45:00.926615Z","shell.execute_reply.started":"2024-08-01T00:44:58.286097Z","shell.execute_reply":"2024-08-01T00:45:00.925714Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission.to_csv('cnn_mnist_submission3.csv', index=False)\n# print(\"Submission file created.\")\n\n\nif os.path.exists('cnn_mnist_submission3.csv'):\n    print(\"File created successfully.\")\nelse:\n    print(\"File not found.\")","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:45:00.927905Z","iopub.execute_input":"2024-08-01T00:45:00.928200Z","iopub.status.idle":"2024-08-01T00:45:00.966578Z","shell.execute_reply.started":"2024-08-01T00:45:00.928175Z","shell.execute_reply":"2024-08-01T00:45:00.965707Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"!head cnn_mnist_submission2.csv","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:45:00.967807Z","iopub.execute_input":"2024-08-01T00:45:00.968358Z","iopub.status.idle":"2024-08-01T00:45:02.001878Z","shell.execute_reply.started":"2024-08-01T00:45:00.968326Z","shell.execute_reply":"2024-08-01T00:45:02.000777Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Predictions Visualization & Errors in Validation Data\n----","metadata":{}},{"cell_type":"code","source":"random_idx = np.random.choice(X_test.shape[0], 15, replace=False)\nfor i, idx in enumerate(random_idx):\n    plt.subplot(3, 5, i+1)\n    plt.imshow(X_test[idx, :, :, 0])\n    plt.xticks([])  # Remove coordinates from axes\n    plt.yticks([])\n    plt.title(submission.Label[idx])","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:45:02.003381Z","iopub.execute_input":"2024-08-01T00:45:02.003759Z","iopub.status.idle":"2024-08-01T00:45:02.930660Z","shell.execute_reply.started":"2024-08-01T00:45:02.003727Z","shell.execute_reply":"2024-08-01T00:45:02.929711Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def display_errors(errors_index,img_errors,pred_errors, obs_errors, input_shape):\n    \"\"\" This function shows 6 images with their predicted and real labels\"\"\"\n    n = 0\n    nrows = 2\n    ncols = 3\n    fig, ax = plt.subplots(nrows,ncols,sharex=True,sharey=True)\n    for row in range(nrows):\n        for col in range(ncols):\n            error = errors_index[n]\n            ax[row,col].imshow((img_errors[error]).reshape((input_shape[0], input_shape[1])))\n            ax[row,col].set_title(\"Predicted label :{}\\nTrue label :{}\".format(pred_errors[error],obs_errors[error]))\n            n += 1","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:45:02.932051Z","iopub.execute_input":"2024-08-01T00:45:02.932729Z","iopub.status.idle":"2024-08-01T00:45:02.940140Z","shell.execute_reply.started":"2024-08-01T00:45:02.932686Z","shell.execute_reply":"2024-08-01T00:45:02.939175Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Predict the values from the validation dataset\nY_pred = model.predict(X_val)\n\n# Convert predictions classes to one hot vectors \nY_pred_classes = np.argmax(Y_pred,axis = 1) \n\n# Convert validation observations to one hot vectors\nY_true = np.argmax(Y_val,axis = 1)\n\n#---------------------------------------------------------------------------------------\n# Show some wrong results, and the difference between the predicted label and the real labe\nerrors = (Y_pred_classes - Y_true != 0)\n\nY_pred_classes_errors = Y_pred_classes[errors]\nY_pred_errors = Y_pred[errors]\nY_true_errors = Y_true[errors]\nX_val_errors = X_val[errors]\n\n#---------------------------------------------------------------------------------------\n# Probabilities of the wrong predicted numbers\nY_pred_errors_prob = np.max(Y_pred_errors,axis = 1)\n\n# Predicted probabilities of the true values in the error set\ntrue_prob_errors = np.diagonal(np.take(Y_pred_errors, Y_true_errors, axis=1))\n\n# Difference between the probability of the predicted label and the true label\ndelta_pred_true_errors = Y_pred_errors_prob - true_prob_errors\n\n# Sorted list of the delta prob errors\nsorted_delta_errors = np.argsort(delta_pred_true_errors)\n\n# Top 6 errors \nmost_important_errors = sorted_delta_errors[-6:]\n\n# Show the top 6 errors\ndisplay_errors(most_important_errors, X_val_errors, Y_pred_classes_errors, Y_true_errors, input_shape)","metadata":{"execution":{"iopub.status.busy":"2024-08-01T00:45:02.941187Z","iopub.execute_input":"2024-08-01T00:45:02.941456Z","iopub.status.idle":"2024-08-01T00:45:05.060298Z","shell.execute_reply.started":"2024-08-01T00:45:02.941433Z","shell.execute_reply":"2024-08-01T00:45:05.059397Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}