{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.14","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"gpu","dataSources":[{"sourceId":3362,"databundleVersionId":31148,"sourceType":"competition"}],"dockerImageVersionId":30787,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# 1. Import Packages","metadata":{}},{"cell_type":"code","source":"# Basic\nimport os\nfrom os import listdir\nimport numpy as np\nimport pandas as pd\nimport zipfile\n\n# visuals\nimport matplotlib.pyplot as plt\nfrom matplotlib.image import imread\nfrom PIL import Image\nimport cv2\n\n# Scikit-learn\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import roc_curve, auc, classification_report,confusion_matrix,ConfusionMatrixDisplay\n\n# PyTorch\nimport torch\nimport torch.nn as nn\nimport torch.optim as optim\nfrom torch.optim.lr_scheduler import ReduceLROnPlateau\nfrom torchvision import transforms, models, datasets\nfrom torch.utils.data import Dataset, DataLoader","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2024-11-18T12:54:19.621397Z","iopub.execute_input":"2024-11-18T12:54:19.62224Z","iopub.status.idle":"2024-11-18T12:54:19.628592Z","shell.execute_reply.started":"2024-11-18T12:54:19.622201Z","shell.execute_reply":"2024-11-18T12:54:19.627629Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\nprint(device)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:56:09.783877Z","iopub.execute_input":"2024-11-18T12:56:09.784637Z","iopub.status.idle":"2024-11-18T12:56:09.816573Z","shell.execute_reply.started":"2024-11-18T12:56:09.784597Z","shell.execute_reply":"2024-11-18T12:56:09.815433Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 2. Data","metadata":{}},{"cell_type":"markdown","source":"# 2.1 (Optional) Data unzipping","metadata":{}},{"cell_type":"code","source":"train_path = \"/kaggle/input/dogs-vs-cats/train.zip\"\ntest_path = \"/kaggle/input/dogs-vs-cats/test1.zip\"\n\nfiles = \"/kaggle/working/\"\n\nwith zipfile.ZipFile(train_path, 'r') as zipp:\n    zipp.extractall(files)\n    \nwith zipfile.ZipFile(test_path, 'r') as zipp:\n    zipp.extractall(files)","metadata":{"execution":{"iopub.status.busy":"2024-11-18T12:27:30.376447Z","iopub.execute_input":"2024-11-18T12:27:30.377089Z","iopub.status.idle":"2024-11-18T12:27:49.641327Z","shell.execute_reply.started":"2024-11-18T12:27:30.377052Z","shell.execute_reply":"2024-11-18T12:27:49.640336Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 2.2 (Optional) Label Check","metadata":{}},{"cell_type":"code","source":"image_dir = \"/kaggle/working/train\"\n\nfilenames = os.listdir(image_dir)\n#filenames = sorted(filenames)\n\n!pip install natsort\nfrom natsort import natsorted\nfilenames = natsorted(filenames)\n\nlabels = [x.split(\".\")[0]   for x in filenames]\n\ndata = pd.DataFrame({\"filename\": filenames, \"label\":labels})\n\ndisplay(data.head(10))\ndisplay(data.tail(10))","metadata":{"execution":{"iopub.status.busy":"2024-11-18T12:28:43.855955Z","iopub.execute_input":"2024-11-18T12:28:43.856803Z","iopub.status.idle":"2024-11-18T12:28:56.702806Z","shell.execute_reply.started":"2024-11-18T12:28:43.85676Z","shell.execute_reply":"2024-11-18T12:28:56.701771Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 2.3 EDA","metadata":{}},{"cell_type":"markdown","source":"## 2.3.1. Visulization","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(10,10))\n\nfor i in range(25):\n    plt.subplot(5,5,i+1)\n    filename = 'train/' + 'dog.' + str(i) + '.jpg'\n    image = imread(filename)\n    plt.imshow(image)\n    plt.title('Dog')\n    plt.axis('off')","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:29:02.354954Z","iopub.execute_input":"2024-11-18T12:29:02.355344Z","iopub.status.idle":"2024-11-18T12:29:04.689852Z","shell.execute_reply.started":"2024-11-18T12:29:02.355309Z","shell.execute_reply":"2024-11-18T12:29:04.688881Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"plt.figure(figsize=(10,10))\n\nfor i in range(25):\n    plt.subplot(5,5,i+1)\n    filename = 'train/' + 'cat.' + str(i) + '.jpg'\n    image = imread(filename)\n    plt.imshow(image)\n    plt.title('Cat')\n    plt.axis('off')","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:29:09.184525Z","iopub.execute_input":"2024-11-18T12:29:09.184874Z","iopub.status.idle":"2024-11-18T12:29:11.690995Z","shell.execute_reply.started":"2024-11-18T12:29:09.184842Z","shell.execute_reply":"2024-11-18T12:29:11.690123Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"## 2.3.2. Label Balance Check","metadata":{}},{"cell_type":"code","source":"label_counts = data['label'].value_counts()\n\nprint(label_counts)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:29:48.855115Z","iopub.execute_input":"2024-11-18T12:29:48.855765Z","iopub.status.idle":"2024-11-18T12:29:48.873466Z","shell.execute_reply.started":"2024-11-18T12:29:48.855726Z","shell.execute_reply":"2024-11-18T12:29:48.872429Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 2.3.3. Image Size Check","metadata":{}},{"cell_type":"code","source":"def get_image_size(filepath):\n    with Image.open(filepath) as img:\n        width, height = img.size\n    return width, height\n \ndata['width'], data['height'] = zip(*data['filename'].apply(lambda x: get_image_size(os.path.join(image_dir, x))))\n\ndata.head(10)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:29:52.045486Z","iopub.execute_input":"2024-11-18T12:29:52.046141Z","iopub.status.idle":"2024-11-18T12:29:55.464126Z","shell.execute_reply.started":"2024-11-18T12:29:52.046102Z","shell.execute_reply":"2024-11-18T12:29:55.463041Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"Image shape\n- NumPy: (height, width, channels) for RGB\n- OpenCV: (height, width, channels) for BGR\n- Pillow: (width, height)\n- Matplotlib: (height, width, channels) for RGB\n- TensorFlow: (height, width, channels) for RGB\n- PyTorch: (channels, height, width) for RGB","metadata":{}},{"cell_type":"code","source":"data['width'].describe()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:30:01.087893Z","iopub.execute_input":"2024-11-18T12:30:01.088296Z","iopub.status.idle":"2024-11-18T12:30:01.101998Z","shell.execute_reply.started":"2024-11-18T12:30:01.08826Z","shell.execute_reply":"2024-11-18T12:30:01.100949Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"data['height'].describe()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:30:02.785396Z","iopub.execute_input":"2024-11-18T12:30:02.785754Z","iopub.status.idle":"2024-11-18T12:30:02.796942Z","shell.execute_reply.started":"2024-11-18T12:30:02.785722Z","shell.execute_reply":"2024-11-18T12:30:02.796003Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"plt.hist(data['width'], bins=20, edgecolor='black')\nplt.xlabel('Image Width (pixels)')\nplt.ylabel('Frequency')\nplt.title('Distribution of Image Widths')\nplt.show()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:30:04.745638Z","iopub.execute_input":"2024-11-18T12:30:04.746023Z","iopub.status.idle":"2024-11-18T12:30:05.01979Z","shell.execute_reply.started":"2024-11-18T12:30:04.745984Z","shell.execute_reply":"2024-11-18T12:30:05.01885Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"plt.hist(data['height'], bins=20, edgecolor='black')\nplt.xlabel('Image Height (pixels)')\nplt.ylabel('Frequency')\nplt.title('Distribution of Image Heights')\nplt.show()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:30:07.531401Z","iopub.execute_input":"2024-11-18T12:30:07.532152Z","iopub.status.idle":"2024-11-18T12:30:07.844723Z","shell.execute_reply.started":"2024-11-18T12:30:07.532112Z","shell.execute_reply":"2024-11-18T12:30:07.843745Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"data = data.drop(columns=['width'])\ndata = data.drop(columns=['height'])\n\ndata.head()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:30:11.12343Z","iopub.execute_input":"2024-11-18T12:30:11.124248Z","iopub.status.idle":"2024-11-18T12:30:11.142815Z","shell.execute_reply.started":"2024-11-18T12:30:11.124206Z","shell.execute_reply":"2024-11-18T12:30:11.141807Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 2.4 Data Split","metadata":{}},{"cell_type":"code","source":"X_train, X_test = train_test_split(data, test_size=0.2, stratify=data['label'], random_state=99)\nX_train, X_valid = train_test_split(X_train, test_size=0.2, stratify=X_train['label'], random_state=99)\n\nprint('The shape of train data:', X_train.shape)\nprint('The shape of valid data:', X_valid.shape)\nprint('The shape of test data:', X_test.shape)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:30:13.748572Z","iopub.execute_input":"2024-11-18T12:30:13.749436Z","iopub.status.idle":"2024-11-18T12:30:13.835751Z","shell.execute_reply.started":"2024-11-18T12:30:13.749395Z","shell.execute_reply":"2024-11-18T12:30:13.834692Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 2.5 Image Data Generator for Augmentation","metadata":{}},{"cell_type":"code","source":"image_size = 224\nimage_channel = 3\nbatch_size = 64","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:30:18.19879Z","iopub.execute_input":"2024-11-18T12:30:18.199182Z","iopub.status.idle":"2024-11-18T12:30:18.203644Z","shell.execute_reply.started":"2024-11-18T12:30:18.199144Z","shell.execute_reply":"2024-11-18T12:30:18.202675Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"train_transform = transforms.Compose([\n                  transforms.Resize((224,224)),\n                  transforms.RandomHorizontalFlip(),\n                  transforms.RandomAffine(degrees=5, shear=0.1, translate=(0.1, 0.1)),\n                  transforms.RandomResizedCrop(224, scale=(0.8, 1.2)),\n                  transforms.ToTensor(),\n    \n])\n\nvalid_transform = transforms.Compose([\n                  transforms.Resize((224,224)),\n                  transforms.ToTensor(),\n])\n\ntest_transform = transforms.Compose([\n                  transforms.Resize((224,224)),\n                  transforms.ToTensor(),\n])","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T12:39:18.938694Z","iopub.execute_input":"2024-11-18T12:39:18.939709Z","iopub.status.idle":"2024-11-18T12:39:18.947042Z","shell.execute_reply.started":"2024-11-18T12:39:18.939667Z","shell.execute_reply":"2024-11-18T12:39:18.946026Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"![image.png](attachment:515908bd-646e-4b81-b08e-24d2558b16af.png)","metadata":{},"attachments":{"515908bd-646e-4b81-b08e-24d2558b16af.png":{"image/png":"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"}}},{"cell_type":"code","source":"class CustomDataset(Dataset):\n    def __init__(self, dataframe, directory, x_col, y_col, transform=None):\n        self.dataframe = dataframe\n        self.directory = directory\n        self.x_col = x_col\n        self.y_col = y_col\n        self.transform = transform\n        self.label_map = {'cat': 0, 'dog': 1}\n\n    def __len__(self):\n        return len(self.dataframe)\n\n    def __getitem__(self, idx):\n        img_name = os.path.join(self.directory, self.dataframe.iloc[idx][self.x_col])\n        label = self.dataframe.iloc[idx][self.y_col]\n        label = self.label_map[label]\n\n        image = Image.open(img_name).convert('RGB')\n\n        if self.transform:\n            image = self.transform(image)\n\n        label = torch.tensor(label, dtype=torch.float32)\n\n        return image, label\n\ntrain_dataset = CustomDataset(X_train, directory='train/', x_col='filename', y_col='label', transform=train_transform)\nvalid_dataset = CustomDataset(X_valid, directory='train/', x_col='filename', y_col='label', transform=valid_transform)\ntest_dataset = CustomDataset(X_test, directory='train/', x_col='filename', y_col='label', transform=test_transform)\n\ntrain_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True)\nvalid_loader = DataLoader(valid_dataset, batch_size=batch_size, shuffle=False)\ntest_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T14:24:52.538127Z","iopub.execute_input":"2024-11-18T14:24:52.538509Z","iopub.status.idle":"2024-11-18T14:24:52.549225Z","shell.execute_reply.started":"2024-11-18T14:24:52.538473Z","shell.execute_reply":"2024-11-18T14:24:52.548244Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"flow_from_directory\n- Loads images from a directory structure where images are organized into subdirectories based on their classes.\n- main_directory/\n    \n    class_a/\n        img1.jpg\n        img2.jpg\n    class_b/\n        img3.jpg\n        img4.jpg\n\n- Example code:\n\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator\n\ndatagen = ImageDataGenerator(rescale=1.0/255)\n\ntrain_data = datagen.flow_from_directory(\n\n    directory='path/to/main_directory',\n    \n    target_size=(150, 150),\n    \n    batch_size=32,\n    \n    class_mode='binary'  # or 'categorical' for multi-class\n)","metadata":{}},{"cell_type":"markdown","source":"flow_from_dataframe\n- Loads images from file paths listed in a DataFrame. It is useful if images are not organized by class in subdirectories or if additional metadata is associated with the images.\n- | filepath            | label |\n\n| path/to/img1.jpg    | 0     |\n\n| path/to/img2.jpg    | 1     |\n\n- Example code:\n\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator\n\ndatagen = ImageDataGenerator(rescale=1.0/255)\n\ntrain_data = datagen.flow_from_dataframe(\n\n    dataframe=df,\n    \n    x_col='filepath',\n    \n    y_col='label',\n    \n    target_size=(150, 150),\n    \n    batch_size=32,\n    \n    class_mode='binary'  # or 'categorical' for multi-class\n    \n)","metadata":{}},{"cell_type":"markdown","source":"# 3. Model","metadata":{}},{"cell_type":"code","source":"class CustomModel(nn.Module):\n    def __init__(self):\n        super(CustomModel, self).__init__()\n        self.conv1 = nn.Conv2d(in_channels=image_channel, out_channels=32, kernel_size=3, padding=1)\n        self.bn1 = nn.BatchNorm2d(32)\n        self.pool1 = nn.MaxPool2d(kernel_size=2, stride=2)\n        self.relu1 = nn.ReLU()\n        self.dropout1 = nn.Dropout(0.2)\n\n        self.conv2 = nn.Conv2d(32, 64, kernel_size=3, padding=1)\n        self.bn2 = nn.BatchNorm2d(64)\n        self.pool2 = nn.MaxPool2d(kernel_size=2, stride=2)\n        self.relu2 = nn.ReLU()\n        self.dropout2 = nn.Dropout(0.2)\n\n        self.conv3 = nn.Conv2d(64, 128, kernel_size=3, padding=1)\n        self.bn3 = nn.BatchNorm2d(128)\n        self.pool3 = nn.MaxPool2d(kernel_size=2, stride=2)\n        self.relu3 = nn.ReLU()\n        self.dropout3 = nn.Dropout(0.2)\n\n        self.conv4 = nn.Conv2d(128, 256, kernel_size=3, padding=1)\n        self.bn4 = nn.BatchNorm2d(256)\n        self.pool4 = nn.MaxPool2d(kernel_size=2, stride=2)\n        self.relu4 = nn.ReLU()\n        self.dropout4 = nn.Dropout(0.2)\n\n        self.global_avg_pool = nn.AdaptiveAvgPool2d((1,1))\n        self.fc1 = nn.Linear(256, 128)\n        self.bn5 = nn.BatchNorm1d(128)\n        self.relu5 = nn.ReLU()\n        self.dropout5 = nn.Dropout(0.2)\n\n        self.fc2 = nn.Linear(128, 64)\n        self.bn6 = nn.BatchNorm1d(64)\n        self.relu6 = nn.ReLU()\n        self.dropout6 = nn.Dropout(0.2)\n\n        self.output = nn.Linear(64, 1)\n        self.sigmoid = nn.Sigmoid()\n\n    def forward(self, x):\n        x = self.pool1(self.relu1(self.bn1(self.conv1(x))))\n        x = self.dropout1(x)\n\n        x = self.pool2(self.relu2(self.bn2(self.conv2(x))))\n        x = self.dropout2(x)\n\n        x = self.pool3(self.relu3(self.bn3(self.conv3(x))))\n        x = self.dropout3(x)\n\n        x = self.pool4(self.relu4(self.bn4(self.conv4(x))))\n        x = self.dropout4(x)\n\n        x = self.global_avg_pool(x)\n        x = torch.flatten(x, 1)\n        x = self.relu5(self.bn5(self.fc1(x)))\n        x = self.dropout5(x)\n\n        x = self.relu6(self.bn6(self.fc2(x)))\n        x = self.dropout6(x)\n\n        x = self.output(x)\n        x = self.sigmoid(x)\n        \n        return x\n\nmodel = CustomModel().to(device)\n\nprint(model)\n","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T14:24:56.385777Z","iopub.execute_input":"2024-11-18T14:24:56.386602Z","iopub.status.idle":"2024-11-18T14:24:56.412276Z","shell.execute_reply.started":"2024-11-18T14:24:56.38656Z","shell.execute_reply":"2024-11-18T14:24:56.411602Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 4.1 Model Compile","metadata":{}},{"cell_type":"code","source":"criterion = nn.BCELoss()\noptimizer = optim.Adam(model.parameters(), lr=0.01)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T14:25:00.349492Z","iopub.execute_input":"2024-11-18T14:25:00.349862Z","iopub.status.idle":"2024-11-18T14:25:00.354912Z","shell.execute_reply.started":"2024-11-18T14:25:00.349825Z","shell.execute_reply":"2024-11-18T14:25:00.353865Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 4.2 Callbacks","metadata":{}},{"cell_type":"code","source":"scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode='min', factor=0.5, patience=5, min_lr=0.00001)\n\nearly_stopping_patience = 10\nearly_stopping_counter = 0\nbest_val_loss = float('inf')\nbest_model_state = None\n\ncheckpoint_path = 'best_model.pth'","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T14:25:02.28178Z","iopub.execute_input":"2024-11-18T14:25:02.28284Z","iopub.status.idle":"2024-11-18T14:25:02.288708Z","shell.execute_reply.started":"2024-11-18T14:25:02.282784Z","shell.execute_reply":"2024-11-18T14:25:02.287722Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 4.3 Model Fit","metadata":{}},{"cell_type":"code","source":"train_losses = []\nvalid_losses = []\ntrain_accuracies = []\nvalid_accuracies = []\n\nnum_epochs = 100\n\nfor epoch in range(num_epochs):\n    model.train()\n    train_loss = 0.0\n    train_correct = 0\n    train_total = 0\n\n    for step, (images, labels) in enumerate(train_loader):\n        images, labels = images.to(device), labels.to(device)\n\n        optimizer.zero_grad()\n        outputs = model(images)\n        labels = labels.view(-1, 1)\n        loss = criterion(outputs, labels)\n        loss.backward()\n        optimizer.step()\n\n        train_loss = train_loss + loss.item() * images.size(0)\n        predictions = torch.round(outputs)\n        train_correct = train_correct + (predictions==labels).sum().item()\n        train_total = train_total + labels.size(0)\n\n    avg_train_loss = train_loss / train_total\n    train_accuracy = train_correct / train_total\n\n    model.eval()\n    valid_loss = 0.0\n    valid_correct = 0\n    valid_total = 0\n    with torch.no_grad():\n        for step, (images, labels) in enumerate(valid_loader):\n            images, labels = images.to(device), labels.to(device)\n\n            outputs = model(images)\n            labels = labels.view(-1, 1)\n            loss = criterion(outputs, labels)\n\n            valid_loss = valid_loss + loss.item() * images.size(0)\n            predictions = torch.round(outputs)\n            valid_correct = valid_correct + (predictions==labels).sum().item()\n            valid_total = valid_total + labels.size(0)\n\n    avg_valid_loss = valid_loss / valid_total\n    valid_accuracy = valid_correct / valid_total\n    \n    train_losses.append(avg_train_loss)\n    valid_losses.append(avg_valid_loss)\n    train_accuracies.append(train_accuracy)\n    valid_accuracies.append(valid_accuracy)\n\n    print(f'Epoch: {epoch+1}, Train Loss: {avg_train_loss: .4f}, Train Accuracy: {train_accuracy: .4f}, Valid Loss: {avg_valid_loss: .4f}, Valid Accuracy: {valid_accuracy: .4f}')\n\n    scheduler.step(avg_valid_loss)\n\n    if avg_valid_loss < best_val_loss:\n        best_val_loss = avg_train_loss\n        best_model_state = model.state_dict()\n        torch.save(best_model_state, checkpoint_path)\n\n    if avg_valid_loss < best_val_loss:\n        early_stopping_counter = 0\n    else:\n        early_stopping_counter += 1\n        if early_stopping_counter >= early_stopping_patience:\n            print('Early stopping triggered')\n            if best_model_state:\n                model.load_state_dict(best_model_state)\n            break","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T14:25:04.722892Z","iopub.execute_input":"2024-11-18T14:25:04.723308Z","iopub.status.idle":"2024-11-18T16:16:12.09655Z","shell.execute_reply.started":"2024-11-18T14:25:04.723271Z","shell.execute_reply":"2024-11-18T16:16:12.095664Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 5. Model Evaluation","metadata":{}},{"cell_type":"markdown","source":"# 5.1 Analyzing the Training Process","metadata":{}},{"cell_type":"code","source":"plt.figure(figsize=(8, 6))\nplt.title('Cross Entropy Loss',fontsize=15)\nplt.xlabel('Epochs',fontsize=12)\nplt.ylabel('Loss',fontsize=12)\nplt.plot(train_losses, label='Training Loss')\nplt.plot(valid_losses, label='Validation Loss')\nplt.legend(loc='upper right')\nplt.show()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T16:37:46.975951Z","iopub.execute_input":"2024-11-18T16:37:46.976682Z","iopub.status.idle":"2024-11-18T16:37:47.22594Z","shell.execute_reply.started":"2024-11-18T16:37:46.97664Z","shell.execute_reply":"2024-11-18T16:37:47.225025Z"}},"outputs":[],"execution_count":null},{"cell_type":"code","source":"plt.figure(figsize=(8, 6))\nplt.title('Classification Accuracy',fontsize=15)\nplt.xlabel('Epochs',fontsize=12)\nplt.ylabel('Accuracy',fontsize=12)\nplt.plot(train_accuracies, label='Training Accuracy')\nplt.plot(valid_accuracies, label='Validation Accuracy')\nplt.legend(loc='upper left')\nplt.show()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T16:38:30.724503Z","iopub.execute_input":"2024-11-18T16:38:30.724877Z","iopub.status.idle":"2024-11-18T16:38:31.013256Z","shell.execute_reply.started":"2024-11-18T16:38:30.72484Z","shell.execute_reply":"2024-11-18T16:38:31.012406Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 5.2 Test Data Evaluation","metadata":{}},{"cell_type":"code","source":"loaded_model = CustomModel()\nloaded_model.load_state_dict(torch.load('best_model.pth'))\nloaded_model = loaded_model.to(device)\nloaded_model.eval()\n\nmodel.eval()\n\ntest_loss = 0.0\ntest_correct = 0\ntest_total = 0\nwith torch.no_grad():\n    for step, (images, labels) in enumerate(valid_loader):\n        images, labels = images.to(device), labels.to(device)\n        \n        outputs = model(images)\n        labels = labels.view(-1, 1)\n        loss = criterion(outputs, labels)\n        \n        test_loss = test_loss + loss.item() * images.size(0)\n        predictions = torch.round(outputs)\n        test_correct = test_correct + (predictions==labels).sum().item()\n        test_total = test_total + labels.size(0)\n        \navg_test_loss = test_loss / test_total\ntest_accuracy = test_correct / test_total\n\nprint(f\"Test Loss: {test_loss:.4f}\")\nprint(f\"Test Accuracy: {test_accuracy:.2%}\")","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2024-11-18T16:43:29.904385Z","iopub.execute_input":"2024-11-18T16:43:29.905577Z","iopub.status.idle":"2024-11-18T16:43:45.652696Z","shell.execute_reply.started":"2024-11-18T16:43:29.905531Z","shell.execute_reply":"2024-11-18T16:43:45.651731Z"}},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 5.3 ROC Analysis","metadata":{}},{"cell_type":"code","source":"# Step 1: Get true labels and predicted probabilities\ny_true = test_generator.classes  # True binary labels (0 or 1)\ny_pred_probs = loaded_model.predict(test_generator).ravel()  # Predicted probabilities\n\n# Step 2: Compute the ROC curve and AUC\nfpr, tpr, thresholds = roc_curve(y_true, y_pred_probs)\nroc_auc = auc(fpr, tpr)\n\n# Step 3: Plot the ROC curve\nplt.figure(figsize=(8, 6))\nplt.plot(fpr, tpr, color='blue', lw=2, label=f'ROC curve (AUC = {roc_auc:.2f})')\nplt.plot([0, 1], [0, 1], color='gray', lw=2, linestyle='--', label='Chance (AUC = 0.5)')\nplt.xlim([0.0, 1.0])\nplt.ylim([0.0, 1.05])\nplt.xlabel('False Positive Rate')\nplt.ylabel('True Positive Rate')\nplt.title('Receiver Operating Characteristic (ROC) Curve')\nplt.legend(loc='lower right')\nplt.show()","metadata":{"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"# Step 1: Get true labels and predictions\ny_true = test_generator.classes  # True binary labels (0 or 1)\ny_pred_probs = loaded_model.predict(test_generator)  # Predicted probabilities\ny_pred = (y_pred_probs > 0.5).astype(int).reshape(-1)  # Convert probabilities to binary class labels (0 or 1)\n\n# Step 2: Classification Report\nprint(\"Classification Report:\")\nprint(classification_report(y_true, y_pred, target_names=test_generator.class_indices.keys()))\n\n# Step 3: Confusion Matrix\nconf_matrix = confusion_matrix(y_true, y_pred)\nprint(\"Confusion Matrix:\")\nprint(conf_matrix)\n\n# Step 4: Confusion Matrix Display\ndisp = ConfusionMatrixDisplay(confusion_matrix=conf_matrix, display_labels=test_generator.class_indices.keys())\ndisp.plot(cmap=\"Blues\")\nplt.title(\"Confusion Matrix\")\nplt.show()","metadata":{"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"# 5.4 Advanced Analysis (Grad-CAM)","metadata":{}},{"cell_type":"code","source":"for layer in loaded_model.layers:\n    print(layer)","metadata":{"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"# Helper function to preprocess an image to match model input\ndef get_img_array(img_path, size=(224, 224)):\n    img = load_img(img_path, target_size=size)\n    array = img_to_array(img)\n    array = np.expand_dims(array, axis=0)  # Add batch dimension\n    return array / 255.0  # Normalize as in your preprocessing\n\n# Function to generate Grad-CAM heatmap\ndef make_gradcam_heatmap(img_array, model, last_conv_layer_name, pred_index=None):\n    # Create a model that maps the input image to the activations of the last conv layer\n    # as well as the model's output\n    grad_model = Model(\n        inputs=model.inputs, outputs=[model.get_layer(last_conv_layer_name).output, model.output]\n    )\n\n    with tf.GradientTape() as tape:\n        # Forward pass through grad_model\n        conv_outputs, predictions = grad_model(img_array)\n        if pred_index is None:\n            pred_index = tf.argmax(predictions[0])  # Get index of the predicted class\n        class_channel = predictions[:, pred_index]\n\n    # Compute gradients of the top predicted class with respect to the last conv layer\n    grads = tape.gradient(class_channel, conv_outputs)\n\n    # Pool gradients across the width and height dimensions\n    pooled_grads = tf.reduce_mean(grads, axis=(0, 1, 2))\n\n    # Weight the convolution outputs by the pooled gradients\n    conv_outputs = conv_outputs[0]\n    heatmap = conv_outputs @ pooled_grads[..., tf.newaxis]\n    heatmap = tf.squeeze(heatmap)\n\n    # Normalize the heatmap to a range of 0 to 1\n    heatmap = np.maximum(heatmap, 0) / np.max(heatmap)\n    return heatmap  \n\n# Function to overlay Grad-CAM heatmap on the original image\ndef display_gradcam(img_path, heatmap, alpha=0.4):\n    img = cv2.imread(img_path)\n    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n    img = cv2.resize(img, (224, 224))  # Resize image to match heatmap\n\n    # Rescale the heatmap to 0-255 and apply a red-to-blue effect\n    heatmap = np.uint8(255 * heatmap)\n    heatmap = cv2.resize(heatmap, (224, 224))  # Ensure heatmap matches the image size\n    heatmap = cv2.applyColorMap(heatmap, cv2.COLORMAP_COOL)\n\n    # Blend the heatmap with the original image\n    superimposed_img = cv2.addWeighted(img, 1 - alpha, heatmap, alpha, 0)\n\n    plt.imshow(superimposed_img)\n    plt.axis('off')\n    plt.show()","metadata":{"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"# List of image filenames to process\ncat_images = [f'cat.{i}.jpg' for i in range(10)]\ndog_images = [f'dog.{i}.jpg' for i in range(10)]\nimage_filenames = cat_images + dog_images\n\n# Iterate over each image filename, apply Grad-CAM, and display results\nfor filename in image_filenames:\n    img_path = 'train/' + filename  # Full path to the image file\n    \n    # Load and preprocess image for Grad-CAM\n    img_array = get_img_array(img_path, size=(224, 224))  # Ensure size matches model input\n    \n    # Generate Grad-CAM heatmap\n    heatmap = make_gradcam_heatmap(img_array, loaded_model, last_conv_layer_name)\n    \n    # Original image resized to 224x224\n    original_img = cv2.imread(img_path)\n    original_img = cv2.cvtColor(original_img, cv2.COLOR_BGR2RGB)\n    original_img = cv2.resize(original_img, (224, 224))  # Resize to 224x224\n    \n    # Rescale the heatmap to 0-255 and resize to 224x224\n    heatmap = np.uint8(255 * heatmap)\n    heatmap = cv2.resize(heatmap, (224, 224))\n    \n    # Apply red-to-blue colormap and ensure it has 3 channels\n    overlay_img = cv2.applyColorMap(heatmap, cv2.COLORMAP_COOL)\n    \n    # Blend the heatmap with the original image\n    superimposed_img = cv2.addWeighted(original_img, 0.6, overlay_img, 0.4, 0)\n\n    # Plotting all three images side-by-side\n    fig, axes = plt.subplots(1, 3, figsize=(15, 5))\n    axes[0].imshow(original_img)\n    axes[0].set_title(\"Original Image\")\n    axes[0].axis('off')\n    \n    axes[1].imshow(superimposed_img)\n    axes[1].set_title(\"Grad-CAM Overlay\")\n    axes[1].axis('off')\n    \n    axes[2].imshow(overlay_img)\n    axes[2].set_title(\"Standalone Heatmap\")\n    axes[2].axis('off')\n    \n    plt.show()\n","metadata":{"trusted":true},"outputs":[],"execution_count":null}]}