{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"nvidiaTeslaT4","dataSources":[{"sourceId":14774,"databundleVersionId":875431,"sourceType":"competition"},{"sourceId":10464928,"sourceType":"datasetVersion","datasetId":6479069}],"dockerImageVersionId":30823,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import os, sys\nimport cv2\nimport pandas as pd\nfrom PIL import Image\nimport json\nimport math\nimport numpy as np\nimport tensorflow as tf\n\nfrom tensorflow import keras\nfrom tensorflow.keras import layers\nfrom tensorflow.keras.applications import DenseNet169\nfrom tensorflow.keras.callbacks import ModelCheckpoint, Callback, EarlyStopping\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator\nfrom tensorflow.keras.models import Sequential\nfrom keras.optimizers import Adam\nimport matplotlib.pyplot as plt\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import cohen_kappa_score, accuracy_score\nfrom sklearn.metrics import roc_curve, auc\nfrom sklearn.utils import shuffle\nfrom collections import Counter\nimport pickle\nfrom tensorflow.keras.models import Model\nfrom tensorflow.keras.applications.densenet import DenseNet169\nimport scipy\nfrom tqdm import tqdm\nfrom IPython.display import display\nimport warnings\nwarnings.filterwarnings('ignore')\n%matplotlib inline","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:36:41.283714Z","iopub.execute_input":"2025-01-14T04:36:41.284399Z","iopub.status.idle":"2025-01-14T04:36:41.292711Z","shell.execute_reply.started":"2025-01-14T04:36:41.284371Z","shell.execute_reply":"2025-01-14T04:36:41.291746Z"}},"outputs":[],"execution_count":4},{"cell_type":"code","source":"EPOCHS = 10\nBATCH_SIZE = 64\nSEED = 20031976\nLRATE = 0.00005\nVERBOSE=1","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:36:41.293947Z","iopub.execute_input":"2025-01-14T04:36:41.294269Z","iopub.status.idle":"2025-01-14T04:36:41.308369Z","shell.execute_reply.started":"2025-01-14T04:36:41.294238Z","shell.execute_reply":"2025-01-14T04:36:41.307493Z"}},"outputs":[],"execution_count":5},{"cell_type":"code","source":"np.random.seed(SEED)\ntf.random.set_seed(SEED)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:36:41.30984Z","iopub.execute_input":"2025-01-14T04:36:41.310047Z","iopub.status.idle":"2025-01-14T04:36:41.325493Z","shell.execute_reply.started":"2025-01-14T04:36:41.310029Z","shell.execute_reply":"2025-01-14T04:36:41.324445Z"}},"outputs":[],"execution_count":6},{"cell_type":"code","source":"train_df = pd.read_csv('/kaggle/input/aptos2019-blindness-detection/train.csv')\ntest_df = pd.read_csv('/kaggle/input/aptos2019-blindness-detection/test.csv')\nprint(train_df.shape)\nprint(test_df.shape)\ntrain_df.head()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:36:41.326624Z","iopub.execute_input":"2025-01-14T04:36:41.326929Z","iopub.status.idle":"2025-01-14T04:36:41.370132Z","shell.execute_reply.started":"2025-01-14T04:36:41.326902Z","shell.execute_reply":"2025-01-14T04:36:41.36947Z"}},"outputs":[{"name":"stdout","text":"(3662, 2)\n(1928, 1)\n","output_type":"stream"},{"execution_count":7,"output_type":"execute_result","data":{"text/plain":" id_code diagnosis\n0 000c1434d8d7 2\n1 001639a390f0 4\n2 0024cdab0c1e 1\n3 002c21358ce6 0\n4 005b95c28852 0","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
id_codediagnosis
0000c1434d8d72
1001639a390f04
20024cdab0c1e1
3002c21358ce60
4005b95c288520
\n
"},"metadata":{}}],"execution_count":7},{"cell_type":"code","source":"def crop_image(image, tol = 7):\n if image.ndim == 2:\n mask = image > tol\n return image[np.ix_(mask.any(1), mask.any(0))]\n elif image.ndim == 3:\n image_gray = cv2.cvtColor(image, cv2.COLOR_RGB2GRAY)\n mask = image_gray > tol\n check_shape = image[:,:,0][np.ix_(mask.any(1), mask.any(0))].shape[0]\n if(check_shape == 0):\n return image\n else:\n image1 = image [:,:,0][np.ix_(mask.any(1), mask.any(0))]\n image2 = image [:,:,1][np.ix_(mask.any(1), mask.any(0))]\n image3 = image [:,:,2][np.ix_(mask.any(1), mask.any(0))]\n image = np.stack([image1, image2, image3], axis = -1)\n return image\n\ndef add_black_padding_and_resize(image, img_size):\n h, w = image.shape[:2]\n new_h, new_w = img_size, img_size\n if h > w:\n scale_factor = img_size / h\n else:\n scale_factor = img_size / w\n\n new_h = int(h * scale_factor)\n new_w = int(w * scale_factor)\n resized_image = cv2.resize(image, (new_w, new_h))\n top = (img_size - new_h) // 2\n bottom = img_size - new_h - top\n left = (img_size - new_w) // 2\n right = img_size - new_w - left\n\n padded_resized_image = cv2.copyMakeBorder(resized_image, top, bottom, left, right, cv2.BORDER_CONSTANT, value= 0)\n return padded_resized_image","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:36:41.370811Z","iopub.execute_input":"2025-01-14T04:36:41.371045Z","iopub.status.idle":"2025-01-14T04:36:41.378972Z","shell.execute_reply.started":"2025-01-14T04:36:41.371025Z","shell.execute_reply":"2025-01-14T04:36:41.378108Z"}},"outputs":[],"execution_count":8},{"cell_type":"code","source":"def preprocess_image(image, img_size=224):\n image = Image.open(image)\n image = np.array(image)\n image = crop_image(image)\n image = add_black_padding_and_resize(image, img_size)\n image = cv2.cvtColor(image, cv2.COLOR_RGB2GRAY)\n kernel_horizontal = np.array([[1, 1, 1, 1, 1]], dtype=np.uint8)\n\n kernel_vertical = np.array([[1],\n [1],\n [1],\n [1],\n [1]], dtype=np.uint8)\n\n blackhat_h = cv2.morphologyEx(image, cv2.MORPH_BLACKHAT, kernel_horizontal)\n blackhat_v = cv2.morphologyEx(image, cv2.MORPH_BLACKHAT, kernel_vertical)\n kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (9, 9))\n tophat = cv2.morphologyEx(image, cv2.MORPH_TOPHAT, kernel)\n blackhat = cv2.morphologyEx(image, cv2.MORPH_BLACKHAT, kernel)\n image_tophat = cv2.add(image, tophat)\n image_blackhat = np.maximum(np.maximum(blackhat_h, blackhat_v), blackhat)\n image = cv2.subtract(image_tophat, image_blackhat)\n image =np.stack((image,)*3, axis=-1)\n image = Image.fromarray(image)\n return image","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:36:41.37974Z","iopub.execute_input":"2025-01-14T04:36:41.380061Z","iopub.status.idle":"2025-01-14T04:36:41.391436Z","shell.execute_reply.started":"2025-01-14T04:36:41.380028Z","shell.execute_reply":"2025-01-14T04:36:41.390588Z"}},"outputs":[],"execution_count":9},{"cell_type":"code","source":"N = train_df.shape[0]\nx_train = np.empty((N, 224, 224, 3), dtype=np.uint8)\nfor i, image_id in enumerate(tqdm(train_df['id_code'])):\n x_train[i, :, :, :] = preprocess_image(f'/kaggle/input/aptos2019-blindness-detection/train_images/{image_id}.png')","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:36:41.39216Z","iopub.execute_input":"2025-01-14T04:36:41.392347Z","iopub.status.idle":"2025-01-14T04:49:07.476843Z","shell.execute_reply.started":"2025-01-14T04:36:41.39233Z","shell.execute_reply":"2025-01-14T04:49:07.475863Z"}},"outputs":[{"name":"stderr","text":"100%|██████████| 3662/3662 [12:26<00:00, 4.91it/s]\n","output_type":"stream"}],"execution_count":10},{"cell_type":"code","source":"N = test_df.shape[0]\nx_test = np.empty((N, 224, 224, 3), dtype=np.uint8)\nfor i, image_id in enumerate(tqdm(test_df['id_code'])):\n x_test[i, :, :, :] = preprocess_image(f'/kaggle/input/aptos2019-blindness-detection/test_images/{image_id}.png')","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:49:07.479264Z","iopub.execute_input":"2025-01-14T04:49:07.479478Z","iopub.status.idle":"2025-01-14T04:51:20.297819Z","shell.execute_reply.started":"2025-01-14T04:49:07.47946Z","shell.execute_reply":"2025-01-14T04:51:20.296891Z"}},"outputs":[{"name":"stderr","text":"100%|██████████| 1928/1928 [02:12<00:00, 14.52it/s]\n","output_type":"stream"}],"execution_count":11},{"cell_type":"code","source":"y_train = pd.get_dummies(train_df['diagnosis']).values","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:20.299424Z","iopub.execute_input":"2025-01-14T04:51:20.299699Z","iopub.status.idle":"2025-01-14T04:51:20.311938Z","shell.execute_reply.started":"2025-01-14T04:51:20.299669Z","shell.execute_reply":"2025-01-14T04:51:20.311058Z"}},"outputs":[],"execution_count":12},{"cell_type":"code","source":"print(x_train.shape)\nprint(y_train.shape)\nprint(x_test.shape)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:20.312852Z","iopub.execute_input":"2025-01-14T04:51:20.313344Z","iopub.status.idle":"2025-01-14T04:51:20.327294Z","shell.execute_reply.started":"2025-01-14T04:51:20.313311Z","shell.execute_reply":"2025-01-14T04:51:20.326224Z"}},"outputs":[{"name":"stdout","text":"(3662, 224, 224, 3)\n(3662, 5)\n(1928, 224, 224, 3)\n","output_type":"stream"}],"execution_count":13},{"cell_type":"code","source":"from imblearn.over_sampling import SMOTE, ADASYN\nx_resampled, y_resampled = SMOTE(random_state=SEED, sampling_strategy = 'not majority').fit_resample(x_train.reshape(x_train.shape[0], -1), train_df['diagnosis'].ravel())","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:20.328306Z","iopub.execute_input":"2025-01-14T04:51:20.328661Z","iopub.status.idle":"2025-01-14T04:51:29.370683Z","shell.execute_reply.started":"2025-01-14T04:51:20.32863Z","shell.execute_reply":"2025-01-14T04:51:29.369987Z"}},"outputs":[],"execution_count":14},{"cell_type":"code","source":"print(\"x_resampled.shape=\",x_resampled.shape)\nprint(\"y_resampled.shape=\",y_resampled.shape)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:29.371518Z","iopub.execute_input":"2025-01-14T04:51:29.372028Z","iopub.status.idle":"2025-01-14T04:51:29.377129Z","shell.execute_reply.started":"2025-01-14T04:51:29.372005Z","shell.execute_reply":"2025-01-14T04:51:29.376387Z"}},"outputs":[{"name":"stdout","text":"x_resampled.shape= (9025, 150528)\ny_resampled.shape= (9025,)\n","output_type":"stream"}],"execution_count":15},{"cell_type":"code","source":"x_train = x_resampled.reshape(x_resampled.shape[0], 224, 224, 3)\ny_train = pd.get_dummies(y_resampled).values","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:29.378032Z","iopub.execute_input":"2025-01-14T04:51:29.378328Z","iopub.status.idle":"2025-01-14T04:51:29.395775Z","shell.execute_reply.started":"2025-01-14T04:51:29.378296Z","shell.execute_reply":"2025-01-14T04:51:29.394851Z"}},"outputs":[],"execution_count":16},{"cell_type":"code","source":"print(\"x_train.shape=\",x_train.shape)\nprint(\"y_train.shape=\",y_train.shape)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:29.39663Z","iopub.execute_input":"2025-01-14T04:51:29.396945Z","iopub.status.idle":"2025-01-14T04:51:29.407916Z","shell.execute_reply.started":"2025-01-14T04:51:29.396924Z","shell.execute_reply":"2025-01-14T04:51:29.407137Z"}},"outputs":[{"name":"stdout","text":"x_train.shape= (9025, 224, 224, 3)\ny_train.shape= (9025, 5)\n","output_type":"stream"}],"execution_count":17},{"cell_type":"code","source":"y_train_multi = np.empty(y_train.shape, dtype=y_train.dtype)\ny_train_multi[:, 4] = y_train[:, 4]\nfor i in range(3, -1, -1):\n y_train_multi[:, i] = np.logical_or(y_train[:, i], y_train_multi[:, i+1])","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:29.408755Z","iopub.execute_input":"2025-01-14T04:51:29.409056Z","iopub.status.idle":"2025-01-14T04:51:29.419484Z","shell.execute_reply.started":"2025-01-14T04:51:29.409028Z","shell.execute_reply":"2025-01-14T04:51:29.418681Z"}},"outputs":[],"execution_count":18},{"cell_type":"code","source":"print(\"Original y_train:\", y_train.sum(axis=0))\nprint(\"Multilabel version:\", y_train_multi.sum(axis=0))","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:29.420361Z","iopub.execute_input":"2025-01-14T04:51:29.420657Z","iopub.status.idle":"2025-01-14T04:51:29.431704Z","shell.execute_reply.started":"2025-01-14T04:51:29.420624Z","shell.execute_reply":"2025-01-14T04:51:29.430983Z"}},"outputs":[{"name":"stdout","text":"Original y_train: [1805 1805 1805 1805 1805]\nMultilabel version: [9025 7220 5415 3610 1805]\n","output_type":"stream"}],"execution_count":19},{"cell_type":"code","source":"x_sptrain, x_spval, y_sptrain, y_spval = train_test_split(\n x_train, y_train_multi, \n test_size=0.30, \n random_state=SEED\n)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:29.432447Z","iopub.execute_input":"2025-01-14T04:51:29.432712Z","iopub.status.idle":"2025-01-14T04:51:29.823496Z","shell.execute_reply.started":"2025-01-14T04:51:29.432679Z","shell.execute_reply":"2025-01-14T04:51:29.822827Z"}},"outputs":[],"execution_count":20},{"cell_type":"code","source":"def train_datagen():\n return ImageDataGenerator(\n horizontal_flip=True, \n vertical_flip=True, \n rotation_range=60,\n zoom_range=0.15, \n fill_mode='constant',\n cval=0.,\n )","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:29.82417Z","iopub.execute_input":"2025-01-14T04:51:29.824373Z","iopub.status.idle":"2025-01-14T04:51:29.828246Z","shell.execute_reply.started":"2025-01-14T04:51:29.824356Z","shell.execute_reply":"2025-01-14T04:51:29.827398Z"}},"outputs":[],"execution_count":21},{"cell_type":"code","source":"data_generator = train_datagen().flow(x_sptrain, y_sptrain, batch_size=BATCH_SIZE, seed=SEED)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:29.831044Z","iopub.execute_input":"2025-01-14T04:51:29.831279Z","iopub.status.idle":"2025-01-14T04:51:30.733091Z","shell.execute_reply.started":"2025-01-14T04:51:29.831259Z","shell.execute_reply":"2025-01-14T04:51:30.732357Z"}},"outputs":[],"execution_count":22},{"cell_type":"code","source":"import tensorflow as tf\nimport keras.backend as K\n\ndef precision(y_true, y_pred):\n y_true_f = tf.cast(y_true, tf.float32) \n y_pred_f = tf.cast(y_pred, tf.float32) \n true_positives = tf.reduce_sum(tf.round(tf.clip_by_value(y_true_f * y_pred_f, 0, 1)))\n predicted_positives = tf.reduce_sum(tf.round(tf.clip_by_value(y_pred_f, 0, 1)))\n precision = true_positives / (predicted_positives + K.epsilon())\n return precision\n\ndef recall(y_true, y_pred):\n y_true_f = tf.cast(y_true, tf.float32) \n y_pred_f = tf.cast(y_pred, tf.float32) \n true_positives = tf.reduce_sum(tf.round(tf.clip_by_value(y_true_f * y_pred_f, 0, 1)))\n possible_positives = tf.reduce_sum(tf.round(tf.clip_by_value(y_true_f, 0, 1)))\n recall = true_positives / (possible_positives + K.epsilon())\n return recall\n\ndef fbeta_score(y_true, y_pred, beta=1):\n if beta < 0:\n raise ValueError('The lowest choosable beta is zero (only precision).')\n y_true_f = tf.cast(y_true, tf.float32)\n p = precision(y_true_f, y_pred)\n r = recall(y_true_f, y_pred)\n bb = beta ** 2\n\n def fbeta():\n num = (1 + bb) * (p * r)\n den = bb * p + r + K.epsilon()\n return num / den\n\n fbeta_score = tf.cond(\n tf.equal(tf.reduce_sum(tf.round(tf.clip_by_value(y_true_f, 0, 1))), 0),\n lambda: 0.0, \n fbeta \n )\n\n return fbeta_score\n\ndef fmeasure(y_true, y_pred):\n return fbeta_score(y_true, y_pred, beta=1)\n\ndef mean_pred(y_true, y_pred):\n return tf.reduce_mean(tf.cast(y_pred, tf.float32))\n\ndef f1_score(y_true, y_pred):\n p = precision(y_true, y_pred)\n r = recall(y_true, y_pred)\n return 2 * (p * r) / (p + r + K.epsilon())\n\nprint(\"Evaluation metrics defined ...\")","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:30.73436Z","iopub.execute_input":"2025-01-14T04:51:30.734592Z","iopub.status.idle":"2025-01-14T04:51:30.744804Z","shell.execute_reply.started":"2025-01-14T04:51:30.734572Z","shell.execute_reply":"2025-01-14T04:51:30.743919Z"}},"outputs":[{"name":"stdout","text":"Evaluation metrics defined ...\n","output_type":"stream"}],"execution_count":23},{"cell_type":"code","source":"def build_model():\n input_tensor = layers.Input(shape=(224, 224, 3))\n densenet = DenseNet169(\n weights='/kaggle/input/densenet-bc/DenseNet-BC-169-32-no-top.h5',\n include_top=False,\n input_shape=(224,224,3)\n )\n x = densenet(input_tensor)\n x = layers.GlobalAveragePooling2D()(x)\n x = layers.Dropout(0.5)(x)\n output_tensor = layers.Dense(5, activation='sigmoid')(x)\n\n model = Model(inputs=input_tensor, outputs=output_tensor)\n\n model.compile(\n loss='binary_crossentropy',\n optimizer=Adam(learning_rate=LRATE),\n metrics=['accuracy',mean_pred, precision, recall, f1_score, fbeta_score, fmeasure]\n )\n return model","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:30.745652Z","iopub.execute_input":"2025-01-14T04:51:30.745948Z","iopub.status.idle":"2025-01-14T04:51:30.757136Z","shell.execute_reply.started":"2025-01-14T04:51:30.745928Z","shell.execute_reply":"2025-01-14T04:51:30.756227Z"}},"outputs":[],"execution_count":24},{"cell_type":"code","source":"from tensorflow.keras import layers\nmodel = build_model()\nmodel.summary()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:30.75795Z","iopub.execute_input":"2025-01-14T04:51:30.75821Z","iopub.status.idle":"2025-01-14T04:51:36.028633Z","shell.execute_reply.started":"2025-01-14T04:51:30.758183Z","shell.execute_reply":"2025-01-14T04:51:36.027964Z"}},"outputs":[{"output_type":"display_data","data":{"text/plain":"\u001b[1mModel: \"functional\"\u001b[0m\n","text/html":"
Model: \"functional\"\n
\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n│ input_layer (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m224\u001b[0m, \u001b[38;5;34m3\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n│ densenet169 (\u001b[38;5;33mFunctional\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m1664\u001b[0m) │ \u001b[38;5;34m12,642,880\u001b[0m │\n├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n│ global_average_pooling2d │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1664\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n│ (\u001b[38;5;33mGlobalAveragePooling2D\u001b[0m) │ │ │\n├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n│ dropout (\u001b[38;5;33mDropout\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1664\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n│ dense (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m5\u001b[0m) │ \u001b[38;5;34m8,325\u001b[0m │\n└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n","text/html":"
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n┃ Layer (type)                         ┃ Output Shape                ┃         Param # ┃\n┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n│ input_layer (InputLayer)             │ (None, 224, 224, 3)         │               0 │\n├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n│ densenet169 (Functional)             │ (None, 7, 7, 1664)          │      12,642,880 │\n├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n│ global_average_pooling2d             │ (None, 1664)                │               0 │\n│ (GlobalAveragePooling2D)             │                             │                 │\n├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n│ dropout (Dropout)                    │ (None, 1664)                │               0 │\n├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n│ dense (Dense)                        │ (None, 5)                   │           8,325 │\n└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n
\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m12,651,205\u001b[0m (48.26 MB)\n","text/html":"
 Total params: 12,651,205 (48.26 MB)\n
\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m12,492,805\u001b[0m (47.66 MB)\n","text/html":"
 Trainable params: 12,492,805 (47.66 MB)\n
\n"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m158,400\u001b[0m (618.75 KB)\n","text/html":"
 Non-trainable params: 158,400 (618.75 KB)\n
\n"},"metadata":{}}],"execution_count":25},{"cell_type":"code","source":"class KappaMetrics(Callback):\n def __init__(self, validation_data):\n super().__init__()\n self.val_kappas = []\n self.validation_data = validation_data\n\n def on_train_begin(self, logs={}):\n pass\n\n def on_epoch_end(self, epoch, logs={}):\n x_val, y_val = self.validation_data \n y_val = y_val.sum(axis=1) - 1 \n\n y_pred = self.model.predict(x_val) > 0.5\n y_pred = y_pred.astype(int).sum(axis=1) - 1\n\n _val_kappa = cohen_kappa_score(\n y_val,\n y_pred,\n weights='quadratic'\n )\n\n self.val_kappas.append(_val_kappa)\n print(f\"Epoch: {epoch+1} val_kappa: {_val_kappa:.4f}\")\n\n if _val_kappa == max(self.val_kappas):\n print(\"Validation Kappa has improved. Saving model.\")\n self.model.save('/kaggle/working/model.h5') \n\n return","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:36.029387Z","iopub.execute_input":"2025-01-14T04:51:36.029595Z","iopub.status.idle":"2025-01-14T04:51:36.03509Z","shell.execute_reply.started":"2025-01-14T04:51:36.029576Z","shell.execute_reply":"2025-01-14T04:51:36.034172Z"}},"outputs":[],"execution_count":26},{"cell_type":"code","source":"kappa_score = KappaMetrics(validation_data=(x_spval, y_spval))\nhistory = model.fit(\n data_generator,\n steps_per_epoch=x_train.shape[0] // BATCH_SIZE,\n epochs=EPOCHS,\n validation_data=(x_spval, y_spval), \n callbacks=[kappa_score],\n verbose=VERBOSE\n)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T04:51:36.036019Z","iopub.execute_input":"2025-01-14T04:51:36.036329Z","iopub.status.idle":"2025-01-14T07:34:30.51004Z","shell.execute_reply.started":"2025-01-14T04:51:36.036299Z","shell.execute_reply":"2025-01-14T07:34:30.509303Z"}},"outputs":[{"name":"stdout","text":"Epoch 1/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m29s\u001b[0m 214ms/step - accuracy: 0.1844 - f1_score: 0.7139 - fbeta_score: 0.7139 - fm\nEpoch: 1 val_kappa: 0.7369\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m559s\u001b[0m 2s/step - accuracy: 0.2080 - f1_score: 0.7402 - fbeta_score: 0.7402 - fmeasure: 0.7402 - loss: 0.6021 - mean_pred: 0.5821 - precision: 0.7298 - recall: 0.7524 - val_accuracy: 0.2581 - val_f1_score: 0.8881 - val_fbeta_score: 0.8881 - val_fmeasure: 0.8881 - val_loss: 0.3294 - val_mean_pred: 0.5992 - val_precision: 0.8635 - val_recall: 0.9161\nEpoch 2/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.3563 - f1_score: 0.8958 - fbeta_score: 0.8958 - fme\nEpoch: 2 val_kappa: 0.7960\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m95s\u001b[0m 642ms/step - accuracy: 0.3568 - f1_score: 0.8970 - fbeta_score: 0.8970 - fmeasure: 0.8970 - loss: 0.2818 - mean_pred: 0.5923 - precision: 0.9026 - recall: 0.8931 - val_accuracy: 0.2973 - val_f1_score: 0.9136 - val_fbeta_score: 0.9136 - val_fmeasure: 0.9136 - val_loss: 0.2370 - val_mean_pred: 0.5804 - val_precision: 0.9311 - val_recall: 0.8985\nEpoch 3/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.3534 - f1_score: 0.9144 - fbeta_score: 0.9144 - fme\nEpoch: 3 val_kappa: 0.7879\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.3502 - f1_score: 0.9147 - fbeta_score: 0.9147 - fmeasure: 0.9147 - loss: 0.2353 - mean_pred: 0.5963 - precision: 0.9247 - recall: 0.9060 - val_accuracy: 0.2674 - val_f1_score: 0.9123 - val_fbeta_score: 0.9123 - val_fmeasure: 0.9123 - val_loss: 0.2274 - val_mean_pred: 0.5782 - val_precision: 0.9336 - val_recall: 0.8937\nEpoch 4/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.3489 - f1_score: 0.9232 - fbeta_score: 0.9232 - fme\nEpoch: 4 val_kappa: 0.7677\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 627ms/step - accuracy: 0.3497 - f1_score: 0.9232 - fbeta_score: 0.9232 - fmeasure: 0.9232 - loss: 0.2091 - mean_pred: 0.5982 - precision: 0.9303 - recall: 0.9173 - val_accuracy: 0.3936 - val_f1_score: 0.9081 - val_fbeta_score: 0.9081 - val_fmeasure: 0.9081 - val_loss: 0.2343 - val_mean_pred: 0.5455 - val_precision: 0.9575 - val_recall: 0.8647\nEpoch 5/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.3812 - f1_score: 0.9288 - fbeta_score: 0.9288 - fme\nEpoch: 5 val_kappa: 0.7301\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 627ms/step - accuracy: 0.3791 - f1_score: 0.9292 - fbeta_score: 0.9292 - fmeasure: 0.9292 - loss: 0.1958 - mean_pred: 0.5965 - precision: 0.9374 - recall: 0.9216 - val_accuracy: 0.2991 - val_f1_score: 0.9013 - val_fbeta_score: 0.9013 - val_fmeasure: 0.9013 - val_loss: 0.2483 - val_mean_pred: 0.5320 - val_precision: 0.9659 - val_recall: 0.8463\nEpoch 6/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.3819 - f1_score: 0.9310 - fbeta_score: 0.9310 - fme\nEpoch: 6 val_kappa: 0.7886\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.3828 - f1_score: 0.9319 - fbeta_score: 0.9319 - fmeasure: 0.9319 - loss: 0.1882 - mean_pred: 0.6013 - precision: 0.9390 - recall: 0.9256 - val_accuracy: 0.3863 - val_f1_score: 0.9161 - val_fbeta_score: 0.9161 - val_fmeasure: 0.9161 - val_loss: 0.2179 - val_mean_pred: 0.5432 - val_precision: 0.9663 - val_recall: 0.8723\nEpoch 7/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.3831 - f1_score: 0.9362 - fbeta_score: 0.9362 - fme\nEpoch: 7 val_kappa: 0.6426\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.3813 - f1_score: 0.9367 - fbeta_score: 0.9367 - fmeasure: 0.9367 - loss: 0.1754 - mean_pred: 0.6001 - precision: 0.9408 - recall: 0.9334 - val_accuracy: 0.4594 - val_f1_score: 0.8790 - val_fbeta_score: 0.8790 - val_fmeasure: 0.8790 - val_loss: 0.3108 - val_mean_pred: 0.4907 - val_precision: 0.9859 - val_recall: 0.7948\nEpoch 8/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.3787 - f1_score: 0.9415 - fbeta_score: 0.9415 - fme\nEpoch: 8 val_kappa: 0.7213\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.3787 - f1_score: 0.9414 - fbeta_score: 0.9414 - fmeasure: 0.9414 - loss: 0.1653 - mean_pred: 0.6026 - precision: 0.9446 - recall: 0.9391 - val_accuracy: 0.3567 - val_f1_score: 0.9015 - val_fbeta_score: 0.9015 - val_fmeasure: 0.9015 - val_loss: 0.2736 - val_mean_pred: 0.5110 - val_precision: 0.9824 - val_recall: 0.8345\nEpoch 9/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.3872 - f1_score: 0.9438 - fbeta_score: 0.9438 - fme\nEpoch: 9 val_kappa: 0.7773\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.3879 - f1_score: 0.9440 - fbeta_score: 0.9440 - fmeasure: 0.9440 - loss: 0.1564 - mean_pred: 0.6000 - precision: 0.9479 - recall: 0.9409 - val_accuracy: 0.5129 - val_f1_score: 0.9115 - val_fbeta_score: 0.9115 - val_fmeasure: 0.9115 - val_loss: 0.2389 - val_mean_pred: 0.5382 - val_precision: 0.9635 - val_recall: 0.8661\nEpoch 10/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 89ms/stepstep - accuracy: 0.4002 - f1_score: 0.9458 - fbeta_score: 0.9458 - fme\nEpoch: 10 val_kappa: 0.7181\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4003 - f1_score: 0.9459 - fbeta_score: 0.9459 - fmeasure: 0.9459 - loss: 0.1527 - mean_pred: 0.6030 - precision: 0.9474 - recall: 0.9453 - val_accuracy: 0.4823 - val_f1_score: 0.8956 - val_fbeta_score: 0.8956 - val_fmeasure: 0.8956 - val_loss: 0.3018 - val_mean_pred: 0.5070 - val_precision: 0.9793 - val_recall: 0.8267\nEpoch 11/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4042 - f1_score: 0.9490 - fbeta_score: 0.9490 - fme\nEpoch: 11 val_kappa: 0.7007\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4030 - f1_score: 0.9492 - fbeta_score: 0.9492 - fmeasure: 0.9492 - loss: 0.1409 - mean_pred: 0.6019 - precision: 0.9533 - recall: 0.9460 - val_accuracy: 0.3789 - val_f1_score: 0.8948 - val_fbeta_score: 0.8948 - val_fmeasure: 0.8948 - val_loss: 0.3244 - val_mean_pred: 0.5081 - val_precision: 0.9772 - val_recall: 0.8269\nEpoch 12/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4123 - f1_score: 0.9540 - fbeta_score: 0.9540 - fme\nEpoch: 12 val_kappa: 0.7982\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m95s\u001b[0m 639ms/step - accuracy: 0.4171 - f1_score: 0.9535 - fbeta_score: 0.9535 - fmeasure: 0.9535 - loss: 0.1343 - mean_pred: 0.6053 - precision: 0.9535 - recall: 0.9545 - val_accuracy: 0.4886 - val_f1_score: 0.9201 - val_fbeta_score: 0.9201 - val_fmeasure: 0.9201 - val_loss: 0.2276 - val_mean_pred: 0.5386 - val_precision: 0.9729 - val_recall: 0.8738\nEpoch 13/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4456 - f1_score: 0.9574 - fbeta_score: 0.9574 - fme\nEpoch: 13 val_kappa: 0.7949\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4395 - f1_score: 0.9570 - fbeta_score: 0.9570 - fmeasure: 0.9570 - loss: 0.1264 - mean_pred: 0.6035 - precision: 0.9608 - recall: 0.9538 - val_accuracy: 0.3530 - val_f1_score: 0.9203 - val_fbeta_score: 0.9203 - val_fmeasure: 0.9203 - val_loss: 0.2340 - val_mean_pred: 0.5335 - val_precision: 0.9759 - val_recall: 0.8715\nEpoch 14/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4010 - f1_score: 0.9624 - fbeta_score: 0.9624 - fme\nEpoch: 14 val_kappa: 0.6427\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4013 - f1_score: 0.9618 - fbeta_score: 0.9618 - fmeasure: 0.9618 - loss: 0.1142 - mean_pred: 0.6003 - precision: 0.9652 - recall: 0.9592 - val_accuracy: 0.4609 - val_f1_score: 0.8744 - val_fbeta_score: 0.8744 - val_fmeasure: 0.8744 - val_loss: 0.4127 - val_mean_pred: 0.4773 - val_precision: 0.9887 - val_recall: 0.7852\nEpoch 15/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.3944 - f1_score: 0.9618 - fbeta_score: 0.9618 - fme\nEpoch: 15 val_kappa: 0.8201\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m94s\u001b[0m 634ms/step - accuracy: 0.3965 - f1_score: 0.9619 - fbeta_score: 0.9619 - fmeasure: 0.9619 - loss: 0.1114 - mean_pred: 0.6051 - precision: 0.9612 - recall: 0.9631 - val_accuracy: 0.4786 - val_f1_score: 0.9274 - val_fbeta_score: 0.9274 - val_fmeasure: 0.9274 - val_loss: 0.2163 - val_mean_pred: 0.5511 - val_precision: 0.9646 - val_recall: 0.8938\nEpoch 16/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4026 - f1_score: 0.9630 - fbeta_score: 0.9630 - fme\nEpoch: 16 val_kappa: 0.8191\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 627ms/step - accuracy: 0.4034 - f1_score: 0.9633 - fbeta_score: 0.9633 - fmeasure: 0.9633 - loss: 0.1063 - mean_pred: 0.6038 - precision: 0.9649 - recall: 0.9622 - val_accuracy: 0.5849 - val_f1_score: 0.9238 - val_fbeta_score: 0.9238 - val_fmeasure: 0.9238 - val_loss: 0.2171 - val_mean_pred: 0.5374 - val_precision: 0.9749 - val_recall: 0.8788\nEpoch 17/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4129 - f1_score: 0.9662 - fbeta_score: 0.9662 - fme\nEpoch: 17 val_kappa: 0.7917\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4145 - f1_score: 0.9663 - fbeta_score: 0.9663 - fmeasure: 0.9663 - loss: 0.0999 - mean_pred: 0.6014 - precision: 0.9667 - recall: 0.9665 - val_accuracy: 0.6296 - val_f1_score: 0.9206 - val_fbeta_score: 0.9206 - val_fmeasure: 0.9206 - val_loss: 0.2439 - val_mean_pred: 0.5313 - val_precision: 0.9766 - val_recall: 0.8719\nEpoch 18/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4105 - f1_score: 0.9690 - fbeta_score: 0.9690 - fme\nEpoch: 18 val_kappa: 0.7664\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4132 - f1_score: 0.9687 - fbeta_score: 0.9687 - fmeasure: 0.9687 - loss: 0.0937 - mean_pred: 0.6051 - precision: 0.9701 - recall: 0.9678 - val_accuracy: 0.6654 - val_f1_score: 0.9142 - val_fbeta_score: 0.9142 - val_fmeasure: 0.9142 - val_loss: 0.2622 - val_mean_pred: 0.5172 - val_precision: 0.9841 - val_recall: 0.8547\nEpoch 19/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4378 - f1_score: 0.9703 - fbeta_score: 0.9703 - fme\nEpoch: 19 val_kappa: 0.7491\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 623ms/step - accuracy: 0.4399 - f1_score: 0.9706 - fbeta_score: 0.9706 - fmeasure: 0.9706 - loss: 0.0892 - mean_pred: 0.6006 - precision: 0.9720 - recall: 0.9696 - val_accuracy: 0.5731 - val_f1_score: 0.9048 - val_fbeta_score: 0.9048 - val_fmeasure: 0.9048 - val_loss: 0.3184 - val_mean_pred: 0.5059 - val_precision: 0.9859 - val_recall: 0.8372\nEpoch 20/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4297 - f1_score: 0.9725 - fbeta_score: 0.9725 - fme\nEpoch: 20 val_kappa: 0.7314\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4308 - f1_score: 0.9726 - fbeta_score: 0.9726 - fmeasure: 0.9726 - loss: 0.0850 - mean_pred: 0.6039 - precision: 0.9731 - recall: 0.9724 - val_accuracy: 0.5901 - val_f1_score: 0.9042 - val_fbeta_score: 0.9042 - val_fmeasure: 0.9042 - val_loss: 0.3245 - val_mean_pred: 0.5063 - val_precision: 0.9852 - val_recall: 0.8369\nEpoch 21/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4220 - f1_score: 0.9727 - fbeta_score: 0.9727 - fme\nEpoch: 21 val_kappa: 0.7654\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4215 - f1_score: 0.9731 - fbeta_score: 0.9731 - fmeasure: 0.9731 - loss: 0.0811 - mean_pred: 0.6024 - precision: 0.9750 - recall: 0.9715 - val_accuracy: 0.5982 - val_f1_score: 0.9131 - val_fbeta_score: 0.9131 - val_fmeasure: 0.9131 - val_loss: 0.3175 - val_mean_pred: 0.5161 - val_precision: 0.9825 - val_recall: 0.8541\nEpoch 22/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4290 - f1_score: 0.9775 - fbeta_score: 0.9775 - fme\nEpoch: 22 val_kappa: 0.8536\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m95s\u001b[0m 637ms/step - accuracy: 0.4297 - f1_score: 0.9771 - fbeta_score: 0.9771 - fmeasure: 0.9771 - loss: 0.0715 - mean_pred: 0.6030 - precision: 0.9782 - recall: 0.9764 - val_accuracy: 0.5569 - val_f1_score: 0.9371 - val_fbeta_score: 0.9371 - val_fmeasure: 0.9371 - val_loss: 0.2027 - val_mean_pred: 0.5544 - val_precision: 0.9695 - val_recall: 0.9078\nEpoch 23/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.3808 - f1_score: 0.9763 - fbeta_score: 0.9763 - fme\nEpoch: 23 val_kappa: 0.8542\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m94s\u001b[0m 636ms/step - accuracy: 0.3892 - f1_score: 0.9765 - fbeta_score: 0.9765 - fmeasure: 0.9765 - loss: 0.0715 - mean_pred: 0.6066 - precision: 0.9761 - recall: 0.9773 - val_accuracy: 0.6381 - val_f1_score: 0.9393 - val_fbeta_score: 0.9393 - val_fmeasure: 0.9393 - val_loss: 0.1814 - val_mean_pred: 0.5573 - val_precision: 0.9705 - val_recall: 0.9109\nEpoch 24/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4452 - f1_score: 0.9767 - fbeta_score: 0.9767 - fme\nEpoch: 24 val_kappa: 0.8162\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4429 - f1_score: 0.9768 - fbeta_score: 0.9768 - fmeasure: 0.9768 - loss: 0.0683 - mean_pred: 0.6003 - precision: 0.9795 - recall: 0.9746 - val_accuracy: 0.5694 - val_f1_score: 0.9295 - val_fbeta_score: 0.9295 - val_fmeasure: 0.9295 - val_loss: 0.2502 - val_mean_pred: 0.5307 - val_precision: 0.9842 - val_recall: 0.8815\nEpoch 25/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4308 - f1_score: 0.9805 - fbeta_score: 0.9805 - fme\nEpoch: 25 val_kappa: 0.8344\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4281 - f1_score: 0.9800 - fbeta_score: 0.9800 - fmeasure: 0.9800 - loss: 0.0652 - mean_pred: 0.6035 - precision: 0.9823 - recall: 0.9781 - val_accuracy: 0.5306 - val_f1_score: 0.9258 - val_fbeta_score: 0.9258 - val_fmeasure: 0.9258 - val_loss: 0.2660 - val_mean_pred: 0.5575 - val_precision: 0.9552 - val_recall: 0.8993\nEpoch 26/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4044 - f1_score: 0.9787 - fbeta_score: 0.9787 - fme\nEpoch: 26 val_kappa: 0.8275\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4082 - f1_score: 0.9790 - fbeta_score: 0.9790 - fmeasure: 0.9790 - loss: 0.0643 - mean_pred: 0.6020 - precision: 0.9793 - recall: 0.9790 - val_accuracy: 0.4549 - val_f1_score: 0.9332 - val_fbeta_score: 0.9332 - val_fmeasure: 0.9332 - val_loss: 0.2398 - val_mean_pred: 0.5440 - val_precision: 0.9774 - val_recall: 0.8939\nEpoch 27/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4290 - f1_score: 0.9849 - fbeta_score: 0.9849 - fme\nEpoch: 27 val_kappa: 0.7996\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4322 - f1_score: 0.9838 - fbeta_score: 0.9838 - fmeasure: 0.9838 - loss: 0.0537 - mean_pred: 0.6046 - precision: 0.9844 - recall: 0.9834 - val_accuracy: 0.4863 - val_f1_score: 0.9257 - val_fbeta_score: 0.9257 - val_fmeasure: 0.9257 - val_loss: 0.2913 - val_mean_pred: 0.5261 - val_precision: 0.9844 - val_recall: 0.8747\nEpoch 28/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4055 - f1_score: 0.9830 - fbeta_score: 0.9830 - fme\nEpoch: 28 val_kappa: 0.7837\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 623ms/step - accuracy: 0.4103 - f1_score: 0.9830 - fbeta_score: 0.9830 - fmeasure: 0.9830 - loss: 0.0552 - mean_pred: 0.6012 - precision: 0.9824 - recall: 0.9840 - val_accuracy: 0.5820 - val_f1_score: 0.9175 - val_fbeta_score: 0.9175 - val_fmeasure: 0.9175 - val_loss: 0.3251 - val_mean_pred: 0.5181 - val_precision: 0.9851 - val_recall: 0.8597\nEpoch 29/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4357 - f1_score: 0.9834 - fbeta_score: 0.9834 - fme\nEpoch: 29 val_kappa: 0.7562\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m92s\u001b[0m 623ms/step - accuracy: 0.4336 - f1_score: 0.9832 - fbeta_score: 0.9832 - fmeasure: 0.9832 - loss: 0.0526 - mean_pred: 0.6001 - precision: 0.9834 - recall: 0.9834 - val_accuracy: 0.5849 - val_f1_score: 0.9108 - val_fbeta_score: 0.9108 - val_fmeasure: 0.9108 - val_loss: 0.3633 - val_mean_pred: 0.5122 - val_precision: 0.9839 - val_recall: 0.8490\nEpoch 30/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4350 - f1_score: 0.9852 - fbeta_score: 0.9852 - fme\nEpoch: 30 val_kappa: 0.8492\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4365 - f1_score: 0.9851 - fbeta_score: 0.9851 - fmeasure: 0.9851 - loss: 0.0478 - mean_pred: 0.6038 - precision: 0.9849 - recall: 0.9854 - val_accuracy: 0.4476 - val_f1_score: 0.9393 - val_fbeta_score: 0.9393 - val_fmeasure: 0.9393 - val_loss: 0.2453 - val_mean_pred: 0.5491 - val_precision: 0.9771 - val_recall: 0.9055\nEpoch 31/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4430 - f1_score: 0.9871 - fbeta_score: 0.9871 - fme\nEpoch: 31 val_kappa: 0.8739\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m94s\u001b[0m 636ms/step - accuracy: 0.4390 - f1_score: 0.9867 - fbeta_score: 0.9867 - fmeasure: 0.9867 - loss: 0.0426 - mean_pred: 0.6010 - precision: 0.9871 - recall: 0.9863 - val_accuracy: 0.6012 - val_f1_score: 0.9462 - val_fbeta_score: 0.9462 - val_fmeasure: 0.9462 - val_loss: 0.1956 - val_mean_pred: 0.5579 - val_precision: 0.9758 - val_recall: 0.9191\nEpoch 32/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4270 - f1_score: 0.9869 - fbeta_score: 0.9869 - fme\nEpoch: 32 val_kappa: 0.7878\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4289 - f1_score: 0.9866 - fbeta_score: 0.9866 - fmeasure: 0.9866 - loss: 0.0433 - mean_pred: 0.5986 - precision: 0.9871 - recall: 0.9862 - val_accuracy: 0.5569 - val_f1_score: 0.9216 - val_fbeta_score: 0.9216 - val_fmeasure: 0.9216 - val_loss: 0.3560 - val_mean_pred: 0.5267 - val_precision: 0.9805 - val_recall: 0.8706\nEpoch 33/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4366 - f1_score: 0.9874 - fbeta_score: 0.9874 - fme\nEpoch: 33 val_kappa: 0.8526\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4412 - f1_score: 0.9875 - fbeta_score: 0.9875 - fmeasure: 0.9875 - loss: 0.0410 - mean_pred: 0.6004 - precision: 0.9865 - recall: 0.9886 - val_accuracy: 0.6392 - val_f1_score: 0.9396 - val_fbeta_score: 0.9396 - val_fmeasure: 0.9396 - val_loss: 0.2440 - val_mean_pred: 0.5421 - val_precision: 0.9837 - val_recall: 0.9003\nEpoch 34/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4665 - f1_score: 0.9891 - fbeta_score: 0.9891 - fme\nEpoch: 34 val_kappa: 0.8447\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4630 - f1_score: 0.9888 - fbeta_score: 0.9888 - fmeasure: 0.9888 - loss: 0.0375 - mean_pred: 0.6026 - precision: 0.9887 - recall: 0.9891 - val_accuracy: 0.5739 - val_f1_score: 0.9396 - val_fbeta_score: 0.9396 - val_fmeasure: 0.9396 - val_loss: 0.2440 - val_mean_pred: 0.5471 - val_precision: 0.9803 - val_recall: 0.9032\nEpoch 35/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4379 - f1_score: 0.9882 - fbeta_score: 0.9882 - fme\nEpoch: 35 val_kappa: 0.7651\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4370 - f1_score: 0.9880 - fbeta_score: 0.9880 - fmeasure: 0.9880 - loss: 0.0374 - mean_pred: 0.5998 - precision: 0.9879 - recall: 0.9883 - val_accuracy: 0.6935 - val_f1_score: 0.9098 - val_fbeta_score: 0.9098 - val_fmeasure: 0.9098 - val_loss: 0.3688 - val_mean_pred: 0.5186 - val_precision: 0.9764 - val_recall: 0.8530\nEpoch 36/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4497 - f1_score: 0.9893 - fbeta_score: 0.9893 - fme\nEpoch: 36 val_kappa: 0.7903\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4516 - f1_score: 0.9891 - fbeta_score: 0.9891 - fmeasure: 0.9891 - loss: 0.0356 - mean_pred: 0.6035 - precision: 0.9888 - recall: 0.9895 - val_accuracy: 0.6654 - val_f1_score: 0.9227 - val_fbeta_score: 0.9227 - val_fmeasure: 0.9227 - val_loss: 0.3408 - val_mean_pred: 0.5240 - val_precision: 0.9836 - val_recall: 0.8700\nEpoch 37/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4509 - f1_score: 0.9878 - fbeta_score: 0.9878 - fme\nEpoch: 37 val_kappa: 0.7559\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4509 - f1_score: 0.9882 - fbeta_score: 0.9882 - fmeasure: 0.9882 - loss: 0.0378 - mean_pred: 0.6000 - precision: 0.9878 - recall: 0.9889 - val_accuracy: 0.6510 - val_f1_score: 0.9147 - val_fbeta_score: 0.9147 - val_fmeasure: 0.9147 - val_loss: 0.4001 - val_mean_pred: 0.5105 - val_precision: 0.9892 - val_recall: 0.8518\nEpoch 38/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4675 - f1_score: 0.9903 - fbeta_score: 0.9903 - fme\nEpoch: 38 val_kappa: 0.8251\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4631 - f1_score: 0.9902 - fbeta_score: 0.9902 - fmeasure: 0.9902 - loss: 0.0318 - mean_pred: 0.6015 - precision: 0.9908 - recall: 0.9897 - val_accuracy: 0.6917 - val_f1_score: 0.9339 - val_fbeta_score: 0.9339 - val_fmeasure: 0.9339 - val_loss: 0.2817 - val_mean_pred: 0.5367 - val_precision: 0.9834 - val_recall: 0.8900\nEpoch 39/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4564 - f1_score: 0.9903 - fbeta_score: 0.9903 - fme\nEpoch: 39 val_kappa: 0.8344\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4574 - f1_score: 0.9902 - fbeta_score: 0.9902 - fmeasure: 0.9902 - loss: 0.0329 - mean_pred: 0.6016 - precision: 0.9896 - recall: 0.9910 - val_accuracy: 0.7049 - val_f1_score: 0.9341 - val_fbeta_score: 0.9341 - val_fmeasure: 0.9341 - val_loss: 0.2826 - val_mean_pred: 0.5415 - val_precision: 0.9784 - val_recall: 0.8947\nEpoch 40/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4640 - f1_score: 0.9908 - fbeta_score: 0.9908 - fme\nEpoch: 40 val_kappa: 0.7842\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4607 - f1_score: 0.9907 - fbeta_score: 0.9907 - fmeasure: 0.9907 - loss: 0.0320 - mean_pred: 0.6004 - precision: 0.9915 - recall: 0.9899 - val_accuracy: 0.7445 - val_f1_score: 0.9230 - val_fbeta_score: 0.9230 - val_fmeasure: 0.9230 - val_loss: 0.3682 - val_mean_pred: 0.5219 - val_precision: 0.9867 - val_recall: 0.8682\nEpoch 41/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4476 - f1_score: 0.9897 - fbeta_score: 0.9897 - fme\nEpoch: 41 val_kappa: 0.8767\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m94s\u001b[0m 636ms/step - accuracy: 0.4489 - f1_score: 0.9899 - fbeta_score: 0.9899 - fmeasure: 0.9899 - loss: 0.0330 - mean_pred: 0.6012 - precision: 0.9897 - recall: 0.9902 - val_accuracy: 0.5820 - val_f1_score: 0.9494 - val_fbeta_score: 0.9494 - val_fmeasure: 0.9494 - val_loss: 0.2001 - val_mean_pred: 0.5613 - val_precision: 0.9766 - val_recall: 0.9245\nEpoch 42/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4396 - f1_score: 0.9915 - fbeta_score: 0.9915 - fme\nEpoch: 42 val_kappa: 0.8432\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4401 - f1_score: 0.9914 - fbeta_score: 0.9914 - fmeasure: 0.9914 - loss: 0.0278 - mean_pred: 0.6026 - precision: 0.9913 - recall: 0.9915 - val_accuracy: 0.7264 - val_f1_score: 0.9372 - val_fbeta_score: 0.9372 - val_fmeasure: 0.9372 - val_loss: 0.2542 - val_mean_pred: 0.5397 - val_precision: 0.9834 - val_recall: 0.8962\nEpoch 43/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4278 - f1_score: 0.9921 - fbeta_score: 0.9921 - fme\nEpoch: 43 val_kappa: 0.7959\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4320 - f1_score: 0.9919 - fbeta_score: 0.9919 - fmeasure: 0.9919 - loss: 0.0273 - mean_pred: 0.6056 - precision: 0.9915 - recall: 0.9924 - val_accuracy: 0.6976 - val_f1_score: 0.9252 - val_fbeta_score: 0.9252 - val_fmeasure: 0.9252 - val_loss: 0.3245 - val_mean_pred: 0.5258 - val_precision: 0.9856 - val_recall: 0.8728\nEpoch 44/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4356 - f1_score: 0.9905 - fbeta_score: 0.9905 - fme\nEpoch: 44 val_kappa: 0.7462\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4368 - f1_score: 0.9906 - fbeta_score: 0.9906 - fmeasure: 0.9906 - loss: 0.0325 - mean_pred: 0.6025 - precision: 0.9910 - recall: 0.9902 - val_accuracy: 0.7005 - val_f1_score: 0.9112 - val_fbeta_score: 0.9112 - val_fmeasure: 0.9112 - val_loss: 0.4307 - val_mean_pred: 0.5080 - val_precision: 0.9879 - val_recall: 0.8468\nEpoch 45/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4636 - f1_score: 0.9929 - fbeta_score: 0.9929 - fme\nEpoch: 45 val_kappa: 0.8544\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4586 - f1_score: 0.9928 - fbeta_score: 0.9928 - fmeasure: 0.9928 - loss: 0.0249 - mean_pred: 0.6028 - precision: 0.9927 - recall: 0.9930 - val_accuracy: 0.6669 - val_f1_score: 0.9405 - val_fbeta_score: 0.9405 - val_fmeasure: 0.9405 - val_loss: 0.2363 - val_mean_pred: 0.5440 - val_precision: 0.9823 - val_recall: 0.9028\nEpoch 46/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4486 - f1_score: 0.9923 - fbeta_score: 0.9923 - fme\nEpoch: 46 val_kappa: 0.7691\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4518 - f1_score: 0.9922 - fbeta_score: 0.9922 - fmeasure: 0.9922 - loss: 0.0243 - mean_pred: 0.6019 - precision: 0.9912 - recall: 0.9934 - val_accuracy: 0.7744 - val_f1_score: 0.9148 - val_fbeta_score: 0.9148 - val_fmeasure: 0.9148 - val_loss: 0.4271 - val_mean_pred: 0.5151 - val_precision: 0.9845 - val_recall: 0.8557\nEpoch 47/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4679 - f1_score: 0.9931 - fbeta_score: 0.9931 - fme\nEpoch: 47 val_kappa: 0.8607\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4683 - f1_score: 0.9930 - fbeta_score: 0.9930 - fmeasure: 0.9930 - loss: 0.0245 - mean_pred: 0.5967 - precision: 0.9935 - recall: 0.9925 - val_accuracy: 0.6614 - val_f1_score: 0.9430 - val_fbeta_score: 0.9430 - val_fmeasure: 0.9430 - val_loss: 0.2432 - val_mean_pred: 0.5626 - val_precision: 0.9687 - val_recall: 0.9194\nEpoch 48/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4478 - f1_score: 0.9937 - fbeta_score: 0.9937 - fme\nEpoch: 48 val_kappa: 0.8390\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4525 - f1_score: 0.9935 - fbeta_score: 0.9935 - fmeasure: 0.9935 - loss: 0.0219 - mean_pred: 0.6052 - precision: 0.9934 - recall: 0.9936 - val_accuracy: 0.6968 - val_f1_score: 0.9351 - val_fbeta_score: 0.9351 - val_fmeasure: 0.9351 - val_loss: 0.2862 - val_mean_pred: 0.5536 - val_precision: 0.9695 - val_recall: 0.9041\nEpoch 49/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4769 - f1_score: 0.9929 - fbeta_score: 0.9929 - fme\nEpoch: 49 val_kappa: 0.8190\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4730 - f1_score: 0.9930 - fbeta_score: 0.9930 - fmeasure: 0.9930 - loss: 0.0237 - mean_pred: 0.6024 - precision: 0.9931 - recall: 0.9929 - val_accuracy: 0.7049 - val_f1_score: 0.9327 - val_fbeta_score: 0.9327 - val_fmeasure: 0.9327 - val_loss: 0.2961 - val_mean_pred: 0.5347 - val_precision: 0.9839 - val_recall: 0.8877\nEpoch 50/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4480 - f1_score: 0.9941 - fbeta_score: 0.9941 - fme\nEpoch: 50 val_kappa: 0.7722\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4486 - f1_score: 0.9940 - fbeta_score: 0.9940 - fmeasure: 0.9940 - loss: 0.0212 - mean_pred: 0.6053 - precision: 0.9944 - recall: 0.9937 - val_accuracy: 0.6112 - val_f1_score: 0.9177 - val_fbeta_score: 0.9177 - val_fmeasure: 0.9177 - val_loss: 0.3922 - val_mean_pred: 0.5180 - val_precision: 0.9849 - val_recall: 0.8603\nEpoch 51/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4526 - f1_score: 0.9930 - fbeta_score: 0.9930 - fme\nEpoch: 51 val_kappa: 0.8722\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4573 - f1_score: 0.9930 - fbeta_score: 0.9930 - fmeasure: 0.9930 - loss: 0.0220 - mean_pred: 0.6038 - precision: 0.9930 - recall: 0.9932 - val_accuracy: 0.6322 - val_f1_score: 0.9443 - val_fbeta_score: 0.9443 - val_fmeasure: 0.9443 - val_loss: 0.2404 - val_mean_pred: 0.5658 - val_precision: 0.9679 - val_recall: 0.9225\nEpoch 52/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4608 - f1_score: 0.9923 - fbeta_score: 0.9923 - fme\nEpoch: 52 val_kappa: 0.8685\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4642 - f1_score: 0.9925 - fbeta_score: 0.9925 - fmeasure: 0.9925 - loss: 0.0246 - mean_pred: 0.6024 - precision: 0.9924 - recall: 0.9928 - val_accuracy: 0.7212 - val_f1_score: 0.9456 - val_fbeta_score: 0.9456 - val_fmeasure: 0.9456 - val_loss: 0.2353 - val_mean_pred: 0.5517 - val_precision: 0.9811 - val_recall: 0.9135\nEpoch 53/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4776 - f1_score: 0.9927 - fbeta_score: 0.9927 - fme\nEpoch: 53 val_kappa: 0.8518\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4795 - f1_score: 0.9928 - fbeta_score: 0.9928 - fmeasure: 0.9928 - loss: 0.0225 - mean_pred: 0.6025 - precision: 0.9929 - recall: 0.9928 - val_accuracy: 0.7005 - val_f1_score: 0.9405 - val_fbeta_score: 0.9405 - val_fmeasure: 0.9405 - val_loss: 0.2772 - val_mean_pred: 0.5432 - val_precision: 0.9829 - val_recall: 0.9026\nEpoch 54/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4586 - f1_score: 0.9947 - fbeta_score: 0.9947 - fme\nEpoch: 54 val_kappa: 0.8217\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4560 - f1_score: 0.9947 - fbeta_score: 0.9947 - fmeasure: 0.9947 - loss: 0.0190 - mean_pred: 0.5988 - precision: 0.9948 - recall: 0.9946 - val_accuracy: 0.6710 - val_f1_score: 0.9338 - val_fbeta_score: 0.9338 - val_fmeasure: 0.9338 - val_loss: 0.3189 - val_mean_pred: 0.5388 - val_precision: 0.9813 - val_recall: 0.8919\nEpoch 55/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4656 - f1_score: 0.9936 - fbeta_score: 0.9936 - fme\nEpoch: 55 val_kappa: 0.8470\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4667 - f1_score: 0.9936 - fbeta_score: 0.9936 - fmeasure: 0.9936 - loss: 0.0221 - mean_pred: 0.6022 - precision: 0.9926 - recall: 0.9948 - val_accuracy: 0.6433 - val_f1_score: 0.9406 - val_fbeta_score: 0.9406 - val_fmeasure: 0.9406 - val_loss: 0.2889 - val_mean_pred: 0.5481 - val_precision: 0.9799 - val_recall: 0.9054\nEpoch 56/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4547 - f1_score: 0.9949 - fbeta_score: 0.9949 - fme\nEpoch: 56 val_kappa: 0.8173\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4564 - f1_score: 0.9948 - fbeta_score: 0.9948 - fmeasure: 0.9948 - loss: 0.0170 - mean_pred: 0.6014 - precision: 0.9944 - recall: 0.9954 - val_accuracy: 0.6625 - val_f1_score: 0.9333 - val_fbeta_score: 0.9333 - val_fmeasure: 0.9333 - val_loss: 0.3334 - val_mean_pred: 0.5331 - val_precision: 0.9860 - val_recall: 0.8871\nEpoch 57/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4540 - f1_score: 0.9934 - fbeta_score: 0.9934 - fme\nEpoch: 57 val_kappa: 0.8856\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m94s\u001b[0m 634ms/step - accuracy: 0.4510 - f1_score: 0.9931 - fbeta_score: 0.9931 - fmeasure: 0.9931 - loss: 0.0230 - mean_pred: 0.6040 - precision: 0.9933 - recall: 0.9930 - val_accuracy: 0.6233 - val_f1_score: 0.9519 - val_fbeta_score: 0.9519 - val_fmeasure: 0.9519 - val_loss: 0.2013 - val_mean_pred: 0.5691 - val_precision: 0.9722 - val_recall: 0.9333\nEpoch 58/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4461 - f1_score: 0.9937 - fbeta_score: 0.9937 - fme\nEpoch: 58 val_kappa: 0.8610\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4457 - f1_score: 0.9937 - fbeta_score: 0.9937 - fmeasure: 0.9937 - loss: 0.0204 - mean_pred: 0.6046 - precision: 0.9935 - recall: 0.9940 - val_accuracy: 0.6821 - val_f1_score: 0.9438 - val_fbeta_score: 0.9438 - val_fmeasure: 0.9438 - val_loss: 0.2501 - val_mean_pred: 0.5535 - val_precision: 0.9768 - val_recall: 0.9139\nEpoch 59/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4451 - f1_score: 0.9930 - fbeta_score: 0.9930 - fme\nEpoch: 59 val_kappa: 0.8505\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 623ms/step - accuracy: 0.4450 - f1_score: 0.9932 - fbeta_score: 0.9932 - fmeasure: 0.9932 - loss: 0.0228 - mean_pred: 0.5980 - precision: 0.9931 - recall: 0.9935 - val_accuracy: 0.7877 - val_f1_score: 0.9398 - val_fbeta_score: 0.9398 - val_fmeasure: 0.9398 - val_loss: 0.2588 - val_mean_pred: 0.5437 - val_precision: 0.9819 - val_recall: 0.9020\nEpoch 60/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4703 - f1_score: 0.9950 - fbeta_score: 0.9950 - fme\nEpoch: 60 val_kappa: 0.8867\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m94s\u001b[0m 635ms/step - accuracy: 0.4689 - f1_score: 0.9947 - fbeta_score: 0.9947 - fmeasure: 0.9947 - loss: 0.0169 - mean_pred: 0.6024 - precision: 0.9946 - recall: 0.9948 - val_accuracy: 0.6182 - val_f1_score: 0.9538 - val_fbeta_score: 0.9538 - val_fmeasure: 0.9538 - val_loss: 0.2123 - val_mean_pred: 0.5646 - val_precision: 0.9784 - val_recall: 0.9311\nEpoch 61/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4663 - f1_score: 0.9950 - fbeta_score: 0.9950 - fme\nEpoch: 61 val_kappa: 0.8666\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4676 - f1_score: 0.9951 - fbeta_score: 0.9951 - fmeasure: 0.9951 - loss: 0.0172 - mean_pred: 0.6057 - precision: 0.9950 - recall: 0.9953 - val_accuracy: 0.6370 - val_f1_score: 0.9481 - val_fbeta_score: 0.9481 - val_fmeasure: 0.9481 - val_loss: 0.2495 - val_mean_pred: 0.5542 - val_precision: 0.9819 - val_recall: 0.9172\nEpoch 62/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4675 - f1_score: 0.9952 - fbeta_score: 0.9952 - fme\nEpoch: 62 val_kappa: 0.8665\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4686 - f1_score: 0.9952 - fbeta_score: 0.9952 - fmeasure: 0.9952 - loss: 0.0152 - mean_pred: 0.6023 - precision: 0.9953 - recall: 0.9951 - val_accuracy: 0.7053 - val_f1_score: 0.9417 - val_fbeta_score: 0.9417 - val_fmeasure: 0.9417 - val_loss: 0.2551 - val_mean_pred: 0.5779 - val_precision: 0.9543 - val_recall: 0.9300\nEpoch 63/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4627 - f1_score: 0.9950 - fbeta_score: 0.9950 - fme\nEpoch: 63 val_kappa: 0.8602\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4642 - f1_score: 0.9950 - fbeta_score: 0.9950 - fmeasure: 0.9950 - loss: 0.0155 - mean_pred: 0.6024 - precision: 0.9950 - recall: 0.9951 - val_accuracy: 0.6448 - val_f1_score: 0.9453 - val_fbeta_score: 0.9453 - val_fmeasure: 0.9453 - val_loss: 0.2634 - val_mean_pred: 0.5474 - val_precision: 0.9844 - val_recall: 0.9099\nEpoch 64/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4731 - f1_score: 0.9956 - fbeta_score: 0.9956 - fme\nEpoch: 64 val_kappa: 0.8046\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4744 - f1_score: 0.9953 - fbeta_score: 0.9953 - fmeasure: 0.9953 - loss: 0.0161 - mean_pred: 0.6049 - precision: 0.9952 - recall: 0.9955 - val_accuracy: 0.7020 - val_f1_score: 0.9297 - val_fbeta_score: 0.9297 - val_fmeasure: 0.9297 - val_loss: 0.3439 - val_mean_pred: 0.5257 - val_precision: 0.9905 - val_recall: 0.8770\nEpoch 65/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4807 - f1_score: 0.9944 - fbeta_score: 0.9944 - fme\nEpoch: 65 val_kappa: 0.8094\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m92s\u001b[0m 624ms/step - accuracy: 0.4764 - f1_score: 0.9944 - fbeta_score: 0.9944 - fmeasure: 0.9944 - loss: 0.0195 - mean_pred: 0.6012 - precision: 0.9935 - recall: 0.9953 - val_accuracy: 0.6315 - val_f1_score: 0.9311 - val_fbeta_score: 0.9311 - val_fmeasure: 0.9311 - val_loss: 0.3512 - val_mean_pred: 0.5298 - val_precision: 0.9860 - val_recall: 0.8832\nEpoch 66/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4508 - f1_score: 0.9953 - fbeta_score: 0.9953 - fme\nEpoch: 66 val_kappa: 0.7906\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4573 - f1_score: 0.9952 - fbeta_score: 0.9952 - fmeasure: 0.9952 - loss: 0.0148 - mean_pred: 0.6053 - precision: 0.9952 - recall: 0.9953 - val_accuracy: 0.7592 - val_f1_score: 0.9230 - val_fbeta_score: 0.9230 - val_fmeasure: 0.9230 - val_loss: 0.3648 - val_mean_pred: 0.5220 - val_precision: 0.9860 - val_recall: 0.8687\nEpoch 67/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4883 - f1_score: 0.9949 - fbeta_score: 0.9949 - fme\nEpoch: 67 val_kappa: 0.8010\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4850 - f1_score: 0.9949 - fbeta_score: 0.9949 - fmeasure: 0.9949 - loss: 0.0168 - mean_pred: 0.6019 - precision: 0.9953 - recall: 0.9945 - val_accuracy: 0.7278 - val_f1_score: 0.9287 - val_fbeta_score: 0.9287 - val_fmeasure: 0.9287 - val_loss: 0.3801 - val_mean_pred: 0.5292 - val_precision: 0.9844 - val_recall: 0.8801\nEpoch 68/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4779 - f1_score: 0.9965 - fbeta_score: 0.9965 - fme\nEpoch: 68 val_kappa: 0.8616\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4782 - f1_score: 0.9964 - fbeta_score: 0.9964 - fmeasure: 0.9964 - loss: 0.0121 - mean_pred: 0.6005 - precision: 0.9971 - recall: 0.9956 - val_accuracy: 0.7947 - val_f1_score: 0.9445 - val_fbeta_score: 0.9445 - val_fmeasure: 0.9445 - val_loss: 0.2430 - val_mean_pred: 0.5479 - val_precision: 0.9832 - val_recall: 0.9096\nEpoch 69/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4580 - f1_score: 0.9953 - fbeta_score: 0.9953 - fme\nEpoch: 69 val_kappa: 0.8911\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m94s\u001b[0m 635ms/step - accuracy: 0.4595 - f1_score: 0.9954 - fbeta_score: 0.9954 - fmeasure: 0.9954 - loss: 0.0156 - mean_pred: 0.6013 - precision: 0.9944 - recall: 0.9964 - val_accuracy: 0.7149 - val_f1_score: 0.9541 - val_fbeta_score: 0.9541 - val_fmeasure: 0.9541 - val_loss: 0.1903 - val_mean_pred: 0.5668 - val_precision: 0.9770 - val_recall: 0.9329\nEpoch 70/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4776 - f1_score: 0.9961 - fbeta_score: 0.9961 - fme\nEpoch: 70 val_kappa: 0.8305\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4770 - f1_score: 0.9961 - fbeta_score: 0.9961 - fmeasure: 0.9961 - loss: 0.0150 - mean_pred: 0.6040 - precision: 0.9961 - recall: 0.9960 - val_accuracy: 0.6547 - val_f1_score: 0.9355 - val_fbeta_score: 0.9355 - val_fmeasure: 0.9355 - val_loss: 0.3132 - val_mean_pred: 0.5374 - val_precision: 0.9839 - val_recall: 0.8928\nEpoch 71/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.4854 - f1_score: 0.9962 - fbeta_score: 0.9962 - fme\nEpoch: 71 val_kappa: 0.8612\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4862 - f1_score: 0.9962 - fbeta_score: 0.9962 - fmeasure: 0.9962 - loss: 0.0138 - mean_pred: 0.6028 - precision: 0.9960 - recall: 0.9965 - val_accuracy: 0.7614 - val_f1_score: 0.9456 - val_fbeta_score: 0.9456 - val_fmeasure: 0.9456 - val_loss: 0.2551 - val_mean_pred: 0.5484 - val_precision: 0.9827 - val_recall: 0.9119\nEpoch 72/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4689 - f1_score: 0.9961 - fbeta_score: 0.9961 - fme\nEpoch: 72 val_kappa: 0.8498\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4751 - f1_score: 0.9961 - fbeta_score: 0.9961 - fmeasure: 0.9961 - loss: 0.0136 - mean_pred: 0.6081 - precision: 0.9956 - recall: 0.9967 - val_accuracy: 0.7293 - val_f1_score: 0.9409 - val_fbeta_score: 0.9409 - val_fmeasure: 0.9409 - val_loss: 0.2770 - val_mean_pred: 0.5428 - val_precision: 0.9835 - val_recall: 0.9027\nEpoch 73/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5032 - f1_score: 0.9962 - fbeta_score: 0.9962 - fme\nEpoch: 73 val_kappa: 0.8313\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.5001 - f1_score: 0.9961 - fbeta_score: 0.9961 - fmeasure: 0.9961 - loss: 0.0136 - mean_pred: 0.6037 - precision: 0.9964 - recall: 0.9958 - val_accuracy: 0.7371 - val_f1_score: 0.9336 - val_fbeta_score: 0.9336 - val_fmeasure: 0.9336 - val_loss: 0.3097 - val_mean_pred: 0.5373 - val_precision: 0.9829 - val_recall: 0.8902\nEpoch 74/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4900 - f1_score: 0.9961 - fbeta_score: 0.9961 - fme\nEpoch: 74 val_kappa: 0.8714\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.4927 - f1_score: 0.9960 - fbeta_score: 0.9960 - fmeasure: 0.9960 - loss: 0.0136 - mean_pred: 0.6019 - precision: 0.9958 - recall: 0.9962 - val_accuracy: 0.6256 - val_f1_score: 0.9491 - val_fbeta_score: 0.9491 - val_fmeasure: 0.9491 - val_loss: 0.2559 - val_mean_pred: 0.5617 - val_precision: 0.9769 - val_recall: 0.9237\nEpoch 75/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4904 - f1_score: 0.9977 - fbeta_score: 0.9977 - fme\nEpoch: 75 val_kappa: 0.8748\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4914 - f1_score: 0.9976 - fbeta_score: 0.9976 - fmeasure: 0.9976 - loss: 0.0099 - mean_pred: 0.6048 - precision: 0.9977 - recall: 0.9974 - val_accuracy: 0.6832 - val_f1_score: 0.9513 - val_fbeta_score: 0.9513 - val_fmeasure: 0.9513 - val_loss: 0.2387 - val_mean_pred: 0.5618 - val_precision: 0.9780 - val_recall: 0.9269\nEpoch 76/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5189 - f1_score: 0.9952 - fbeta_score: 0.9952 - fme\nEpoch: 76 val_kappa: 0.8707\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.5155 - f1_score: 0.9952 - fbeta_score: 0.9952 - fmeasure: 0.9952 - loss: 0.0155 - mean_pred: 0.5987 - precision: 0.9949 - recall: 0.9955 - val_accuracy: 0.7430 - val_f1_score: 0.9476 - val_fbeta_score: 0.9476 - val_fmeasure: 0.9476 - val_loss: 0.2380 - val_mean_pred: 0.5616 - val_precision: 0.9746 - val_recall: 0.9230\nEpoch 77/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5126 - f1_score: 0.9955 - fbeta_score: 0.9955 - fme\nEpoch: 77 val_kappa: 0.8429\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.5115 - f1_score: 0.9956 - fbeta_score: 0.9956 - fmeasure: 0.9956 - loss: 0.0146 - mean_pred: 0.6020 - precision: 0.9964 - recall: 0.9948 - val_accuracy: 0.6928 - val_f1_score: 0.9401 - val_fbeta_score: 0.9401 - val_fmeasure: 0.9401 - val_loss: 0.2908 - val_mean_pred: 0.5418 - val_precision: 0.9840 - val_recall: 0.9012\nEpoch 78/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4989 - f1_score: 0.9963 - fbeta_score: 0.9963 - fme\nEpoch: 78 val_kappa: 0.8703\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4973 - f1_score: 0.9962 - fbeta_score: 0.9962 - fmeasure: 0.9962 - loss: 0.0131 - mean_pred: 0.6022 - precision: 0.9959 - recall: 0.9965 - val_accuracy: 0.6115 - val_f1_score: 0.9478 - val_fbeta_score: 0.9478 - val_fmeasure: 0.9478 - val_loss: 0.2584 - val_mean_pred: 0.5577 - val_precision: 0.9789 - val_recall: 0.9195\nEpoch 79/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5219 - f1_score: 0.9956 - fbeta_score: 0.9956 - fme\nEpoch: 79 val_kappa: 0.8712\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 627ms/step - accuracy: 0.5174 - f1_score: 0.9956 - fbeta_score: 0.9956 - fmeasure: 0.9956 - loss: 0.0133 - mean_pred: 0.6021 - precision: 0.9958 - recall: 0.9955 - val_accuracy: 0.6082 - val_f1_score: 0.9485 - val_fbeta_score: 0.9485 - val_fmeasure: 0.9485 - val_loss: 0.2569 - val_mean_pred: 0.5586 - val_precision: 0.9788 - val_recall: 0.9209\nEpoch 80/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4954 - f1_score: 0.9968 - fbeta_score: 0.9968 - fme\nEpoch: 80 val_kappa: 0.8287\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.4978 - f1_score: 0.9966 - fbeta_score: 0.9966 - fmeasure: 0.9966 - loss: 0.0117 - mean_pred: 0.6052 - precision: 0.9969 - recall: 0.9963 - val_accuracy: 0.7227 - val_f1_score: 0.9329 - val_fbeta_score: 0.9329 - val_fmeasure: 0.9329 - val_loss: 0.3589 - val_mean_pred: 0.5405 - val_precision: 0.9785 - val_recall: 0.8925\nEpoch 81/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4806 - f1_score: 0.9950 - fbeta_score: 0.9950 - fme\nEpoch: 81 val_kappa: 0.8638\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 628ms/step - accuracy: 0.4820 - f1_score: 0.9951 - fbeta_score: 0.9951 - fmeasure: 0.9951 - loss: 0.0157 - mean_pred: 0.6057 - precision: 0.9955 - recall: 0.9948 - val_accuracy: 0.6795 - val_f1_score: 0.9454 - val_fbeta_score: 0.9454 - val_fmeasure: 0.9454 - val_loss: 0.2721 - val_mean_pred: 0.5525 - val_precision: 0.9813 - val_recall: 0.9129\nEpoch 82/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5210 - f1_score: 0.9966 - fbeta_score: 0.9966 - fme\nEpoch: 82 val_kappa: 0.8152\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.5204 - f1_score: 0.9965 - fbeta_score: 0.9965 - fmeasure: 0.9965 - loss: 0.0117 - mean_pred: 0.6026 - precision: 0.9961 - recall: 0.9969 - val_accuracy: 0.7157 - val_f1_score: 0.9332 - val_fbeta_score: 0.9332 - val_fmeasure: 0.9332 - val_loss: 0.3795 - val_mean_pred: 0.5299 - val_precision: 0.9897 - val_recall: 0.8839\nEpoch 83/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5273 - f1_score: 0.9972 - fbeta_score: 0.9972 - fme\nEpoch: 83 val_kappa: 0.8650\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.5226 - f1_score: 0.9972 - fbeta_score: 0.9972 - fmeasure: 0.9972 - loss: 0.0090 - mean_pred: 0.6005 - precision: 0.9971 - recall: 0.9972 - val_accuracy: 0.7035 - val_f1_score: 0.9471 - val_fbeta_score: 0.9471 - val_fmeasure: 0.9471 - val_loss: 0.2813 - val_mean_pred: 0.5500 - val_precision: 0.9849 - val_recall: 0.9130\nEpoch 84/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.4946 - f1_score: 0.9978 - fbeta_score: 0.9978 - fme\nEpoch: 84 val_kappa: 0.8615\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.4964 - f1_score: 0.9976 - fbeta_score: 0.9976 - fmeasure: 0.9976 - loss: 0.0087 - mean_pred: 0.6048 - precision: 0.9976 - recall: 0.9976 - val_accuracy: 0.6843 - val_f1_score: 0.9442 - val_fbeta_score: 0.9442 - val_fmeasure: 0.9442 - val_loss: 0.2976 - val_mean_pred: 0.5470 - val_precision: 0.9844 - val_recall: 0.9082\nEpoch 85/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5234 - f1_score: 0.9960 - fbeta_score: 0.9960 - fme\nEpoch: 85 val_kappa: 0.8942\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m95s\u001b[0m 637ms/step - accuracy: 0.5196 - f1_score: 0.9959 - fbeta_score: 0.9959 - fmeasure: 0.9959 - loss: 0.0149 - mean_pred: 0.6032 - precision: 0.9956 - recall: 0.9962 - val_accuracy: 0.6182 - val_f1_score: 0.9556 - val_fbeta_score: 0.9556 - val_fmeasure: 0.9556 - val_loss: 0.2311 - val_mean_pred: 0.5719 - val_precision: 0.9742 - val_recall: 0.9383\nEpoch 86/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5247 - f1_score: 0.9961 - fbeta_score: 0.9961 - fme\nEpoch: 86 val_kappa: 0.8867\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.5214 - f1_score: 0.9962 - fbeta_score: 0.9962 - fmeasure: 0.9962 - loss: 0.0126 - mean_pred: 0.6013 - precision: 0.9961 - recall: 0.9963 - val_accuracy: 0.7242 - val_f1_score: 0.9512 - val_fbeta_score: 0.9512 - val_fmeasure: 0.9512 - val_loss: 0.2295 - val_mean_pred: 0.5664 - val_precision: 0.9738 - val_recall: 0.9303\nEpoch 87/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5226 - f1_score: 0.9974 - fbeta_score: 0.9974 - fme\nEpoch: 87 val_kappa: 0.8400\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.5206 - f1_score: 0.9973 - fbeta_score: 0.9973 - fmeasure: 0.9973 - loss: 0.0095 - mean_pred: 0.6014 - precision: 0.9975 - recall: 0.9972 - val_accuracy: 0.7112 - val_f1_score: 0.9373 - val_fbeta_score: 0.9373 - val_fmeasure: 0.9373 - val_loss: 0.3223 - val_mean_pred: 0.5435 - val_precision: 0.9798 - val_recall: 0.8991\nEpoch 88/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.5259 - f1_score: 0.9971 - fbeta_score: 0.9971 - fme\nEpoch: 88 val_kappa: 0.8285\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.5231 - f1_score: 0.9969 - fbeta_score: 0.9969 - fmeasure: 0.9969 - loss: 0.0111 - mean_pred: 0.5976 - precision: 0.9971 - recall: 0.9966 - val_accuracy: 0.6758 - val_f1_score: 0.9344 - val_fbeta_score: 0.9344 - val_fmeasure: 0.9344 - val_loss: 0.4060 - val_mean_pred: 0.5435 - val_precision: 0.9774 - val_recall: 0.8960\nEpoch 89/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.5289 - f1_score: 0.9962 - fbeta_score: 0.9962 - fme\nEpoch: 89 val_kappa: 0.8086\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.5363 - f1_score: 0.9961 - fbeta_score: 0.9961 - fmeasure: 0.9961 - loss: 0.0137 - mean_pred: 0.6041 - precision: 0.9964 - recall: 0.9959 - val_accuracy: 0.7175 - val_f1_score: 0.9274 - val_fbeta_score: 0.9274 - val_fmeasure: 0.9274 - val_loss: 0.3921 - val_mean_pred: 0.5297 - val_precision: 0.9846 - val_recall: 0.8775\nEpoch 90/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5477 - f1_score: 0.9974 - fbeta_score: 0.9974 - fme\nEpoch: 90 val_kappa: 0.8118\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.5456 - f1_score: 0.9973 - fbeta_score: 0.9973 - fmeasure: 0.9973 - loss: 0.0100 - mean_pred: 0.6022 - precision: 0.9973 - recall: 0.9974 - val_accuracy: 0.7692 - val_f1_score: 0.9288 - val_fbeta_score: 0.9288 - val_fmeasure: 0.9288 - val_loss: 0.4460 - val_mean_pred: 0.5309 - val_precision: 0.9831 - val_recall: 0.8809\nEpoch 91/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.5284 - f1_score: 0.9967 - fbeta_score: 0.9967 - fme\nEpoch: 91 val_kappa: 0.8360\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.5325 - f1_score: 0.9967 - fbeta_score: 0.9967 - fmeasure: 0.9967 - loss: 0.0104 - mean_pred: 0.6042 - precision: 0.9967 - recall: 0.9967 - val_accuracy: 0.7389 - val_f1_score: 0.9335 - val_fbeta_score: 0.9335 - val_fmeasure: 0.9335 - val_loss: 0.3813 - val_mean_pred: 0.5368 - val_precision: 0.9835 - val_recall: 0.8892\nEpoch 92/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5408 - f1_score: 0.9969 - fbeta_score: 0.9969 - fme\nEpoch: 92 val_kappa: 0.8461\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.5365 - f1_score: 0.9969 - fbeta_score: 0.9969 - fmeasure: 0.9969 - loss: 0.0099 - mean_pred: 0.6031 - precision: 0.9973 - recall: 0.9965 - val_accuracy: 0.7378 - val_f1_score: 0.9424 - val_fbeta_score: 0.9424 - val_fmeasure: 0.9424 - val_loss: 0.3271 - val_mean_pred: 0.5418 - val_precision: 0.9861 - val_recall: 0.9033\nEpoch 93/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5435 - f1_score: 0.9970 - fbeta_score: 0.9970 - fme\nEpoch: 93 val_kappa: 0.8351\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.5456 - f1_score: 0.9968 - fbeta_score: 0.9968 - fmeasure: 0.9968 - loss: 0.0110 - mean_pred: 0.6052 - precision: 0.9968 - recall: 0.9968 - val_accuracy: 0.7408 - val_f1_score: 0.9323 - val_fbeta_score: 0.9323 - val_fmeasure: 0.9323 - val_loss: 0.3918 - val_mean_pred: 0.5360 - val_precision: 0.9820 - val_recall: 0.8882\nEpoch 94/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5424 - f1_score: 0.9962 - fbeta_score: 0.9962 - fme\nEpoch: 94 val_kappa: 0.7650\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 624ms/step - accuracy: 0.5385 - f1_score: 0.9964 - fbeta_score: 0.9964 - fmeasure: 0.9964 - loss: 0.0116 - mean_pred: 0.6028 - precision: 0.9959 - recall: 0.9968 - val_accuracy: 0.7027 - val_f1_score: 0.9166 - val_fbeta_score: 0.9166 - val_fmeasure: 0.9166 - val_loss: 0.5098 - val_mean_pred: 0.5124 - val_precision: 0.9897 - val_recall: 0.8548\nEpoch 95/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5212 - f1_score: 0.9970 - fbeta_score: 0.9970 - fme\nEpoch: 95 val_kappa: 0.8593\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 628ms/step - accuracy: 0.5258 - f1_score: 0.9968 - fbeta_score: 0.9968 - fmeasure: 0.9968 - loss: 0.0108 - mean_pred: 0.6037 - precision: 0.9969 - recall: 0.9968 - val_accuracy: 0.7869 - val_f1_score: 0.9427 - val_fbeta_score: 0.9427 - val_fmeasure: 0.9427 - val_loss: 0.2635 - val_mean_pred: 0.5525 - val_precision: 0.9780 - val_recall: 0.9107\nEpoch 96/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5306 - f1_score: 0.9959 - fbeta_score: 0.9959 - fme\nEpoch: 96 val_kappa: 0.8443\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.5327 - f1_score: 0.9959 - fbeta_score: 0.9959 - fmeasure: 0.9959 - loss: 0.0134 - mean_pred: 0.6022 - precision: 0.9955 - recall: 0.9963 - val_accuracy: 0.7286 - val_f1_score: 0.9424 - val_fbeta_score: 0.9424 - val_fmeasure: 0.9424 - val_loss: 0.3048 - val_mean_pred: 0.5509 - val_precision: 0.9793 - val_recall: 0.9091\nEpoch 97/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5473 - f1_score: 0.9966 - fbeta_score: 0.9966 - fme\nEpoch: 97 val_kappa: 0.8375\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.5486 - f1_score: 0.9965 - fbeta_score: 0.9965 - fmeasure: 0.9965 - loss: 0.0119 - mean_pred: 0.6024 - precision: 0.9971 - recall: 0.9960 - val_accuracy: 0.7097 - val_f1_score: 0.9382 - val_fbeta_score: 0.9382 - val_fmeasure: 0.9382 - val_loss: 0.2855 - val_mean_pred: 0.5439 - val_precision: 0.9812 - val_recall: 0.8999\nEpoch 98/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 89ms/stepstep - accuracy: 0.5802 - f1_score: 0.9961 - fbeta_score: 0.9961 - fme\nEpoch: 98 val_kappa: 0.9027\nValidation Kappa has improved. Saving model.\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m96s\u001b[0m 646ms/step - accuracy: 0.5743 - f1_score: 0.9959 - fbeta_score: 0.9959 - fmeasure: 0.9959 - loss: 0.0132 - mean_pred: 0.6001 - precision: 0.9961 - recall: 0.9957 - val_accuracy: 0.6802 - val_f1_score: 0.9573 - val_fbeta_score: 0.9573 - val_fmeasure: 0.9573 - val_loss: 0.1795 - val_mean_pred: 0.5826 - val_precision: 0.9659 - val_recall: 0.9494\nEpoch 99/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5578 - f1_score: 0.9957 - fbeta_score: 0.9957 - fme\nEpoch: 99 val_kappa: 0.8278\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 625ms/step - accuracy: 0.5594 - f1_score: 0.9957 - fbeta_score: 0.9957 - fmeasure: 0.9957 - loss: 0.0133 - mean_pred: 0.6019 - precision: 0.9952 - recall: 0.9963 - val_accuracy: 0.7507 - val_f1_score: 0.9354 - val_fbeta_score: 0.9354 - val_fmeasure: 0.9354 - val_loss: 0.3076 - val_mean_pred: 0.5395 - val_precision: 0.9824 - val_recall: 0.8936\nEpoch 100/100\n\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 90ms/stepstep - accuracy: 0.5806 - f1_score: 0.9973 - fbeta_score: 0.9973 - fme\nEpoch: 100 val_kappa: 0.8530\n\u001b[1m141/141\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m93s\u001b[0m 626ms/step - accuracy: 0.5813 - f1_score: 0.9974 - fbeta_score: 0.9974 - fmeasure: 0.9974 - loss: 0.0079 - mean_pred: 0.5996 - precision: 0.9976 - recall: 0.9973 - val_accuracy: 0.7530 - val_f1_score: 0.9418 - val_fbeta_score: 0.9418 - val_fmeasure: 0.9418 - val_loss: 0.3046 - val_mean_pred: 0.5518 - val_precision: 0.9782 - val_recall: 0.9088\n","output_type":"stream"}],"execution_count":27},{"cell_type":"code","source":"with open('history.json', 'w') as f:\n json.dump(history.history, f)\n\nhistory_df = pd.DataFrame(history.history)\nhistory_df.head(EPOCHS)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T07:34:30.511199Z","iopub.execute_input":"2025-01-14T07:34:30.511493Z","iopub.status.idle":"2025-01-14T07:34:30.53798Z","shell.execute_reply.started":"2025-01-14T07:34:30.511463Z","shell.execute_reply":"2025-01-14T07:34:30.53714Z"}},"outputs":[{"execution_count":28,"output_type":"execute_result","data":{"text/plain":" accuracy f1_score fbeta_score fmeasure loss mean_pred precision \\\n0 0.263574 0.802133 0.802133 0.802133 0.487471 0.586234 0.790402 \n1 0.358081 0.900032 0.900032 0.900032 0.272973 0.594397 0.905124 \n2 0.342409 0.915299 0.915299 0.915299 0.231600 0.598233 0.924307 \n3 0.351591 0.923258 0.923258 0.923258 0.209457 0.601440 0.930345 \n4 0.374228 0.929991 0.929991 0.929991 0.195853 0.599353 0.938617 \n.. ... ... ... ... ... ... ... \n95 0.537755 0.995818 0.995818 0.995818 0.013262 0.603082 0.995408 \n96 0.551686 0.996405 0.996405 0.996405 0.012317 0.602377 0.996697 \n97 0.560551 0.995387 0.995387 0.995387 0.014575 0.602461 0.995488 \n98 0.562925 0.995885 0.995885 0.995885 0.012871 0.603063 0.995295 \n99 0.582872 0.997782 0.997782 0.997782 0.007367 0.602183 0.997692 \n\n recall val_accuracy val_f1_score val_fbeta_score val_fmeasure \\\n0 0.815920 0.258124 0.888120 0.888120 0.888120 \n1 0.896548 0.297267 0.913597 0.913597 0.913597 \n2 0.907747 0.267356 0.912302 0.912302 0.912302 \n3 0.917391 0.393648 0.908070 0.908070 0.908070 \n4 0.922240 0.299114 0.901267 0.901267 0.901267 \n.. ... ... ... ... ... \n95 0.996264 0.728582 0.942434 0.942434 0.942434 \n96 0.996144 0.709749 0.938237 0.938237 0.938237 \n97 0.995343 0.680207 0.957275 0.957275 0.957275 \n98 0.996512 0.750739 0.935350 0.935350 0.935350 \n99 0.997888 0.752954 0.941826 0.941826 0.941826 \n\n val_loss val_mean_pred val_precision val_recall \n0 0.329385 0.599165 0.863541 0.916065 \n1 0.236958 0.580379 0.931052 0.898494 \n2 0.227415 0.578246 0.933612 0.893693 \n3 0.234286 0.545488 0.957546 0.864716 \n4 0.248333 0.531994 0.965927 0.846297 \n.. ... ... ... ... \n95 0.304780 0.550942 0.979318 0.909147 \n96 0.285519 0.543945 0.981237 0.899884 \n97 0.179520 0.582554 0.965924 0.949351 \n98 0.307612 0.539516 0.982410 0.893614 \n99 0.304608 0.551813 0.978241 0.908832 \n\n[100 rows x 16 columns]","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
accuracyf1_scorefbeta_scorefmeasurelossmean_predprecisionrecallval_accuracyval_f1_scoreval_fbeta_scoreval_fmeasureval_lossval_mean_predval_precisionval_recall
00.2635740.8021330.8021330.8021330.4874710.5862340.7904020.8159200.2581240.8881200.8881200.8881200.3293850.5991650.8635410.916065
10.3580810.9000320.9000320.9000320.2729730.5943970.9051240.8965480.2972670.9135970.9135970.9135970.2369580.5803790.9310520.898494
20.3424090.9152990.9152990.9152990.2316000.5982330.9243070.9077470.2673560.9123020.9123020.9123020.2274150.5782460.9336120.893693
30.3515910.9232580.9232580.9232580.2094570.6014400.9303450.9173910.3936480.9080700.9080700.9080700.2342860.5454880.9575460.864716
40.3742280.9299910.9299910.9299910.1958530.5993530.9386170.9222400.2991140.9012670.9012670.9012670.2483330.5319940.9659270.846297
...................................................
950.5377550.9958180.9958180.9958180.0132620.6030820.9954080.9962640.7285820.9424340.9424340.9424340.3047800.5509420.9793180.909147
960.5516860.9964050.9964050.9964050.0123170.6023770.9966970.9961440.7097490.9382370.9382370.9382370.2855190.5439450.9812370.899884
970.5605510.9953870.9953870.9953870.0145750.6024610.9954880.9953430.6802070.9572750.9572750.9572750.1795200.5825540.9659240.949351
980.5629250.9958850.9958850.9958850.0128710.6030630.9952950.9965120.7507390.9353500.9353500.9353500.3076120.5395160.9824100.893614
990.5828720.9977820.9977820.9977820.0073670.6021830.9976920.9978880.7529540.9418260.9418260.9418260.3046080.5518130.9782410.908832
\n

100 rows × 16 columns

\n
"},"metadata":{}}],"execution_count":28},{"cell_type":"code","source":"f1, (ax1, ax2, ax3, ax4) = plt.subplots(1, 4, figsize=(24, 4))\nt1 = f1.suptitle('CNN Performance', fontsize=12)\nf1.subplots_adjust(top=0.85, wspace=0.3)\n\nepoch_list = list(range(1,EPOCHS + 1))\nax1.plot(epoch_list, history.history['accuracy'], label='Train Accuracy')\nax1.plot(epoch_list, history.history['val_accuracy'], label='Validation Accuracy')\nax1.set_xticks(np.arange(0, EPOCHS + 1, 5))\nax1.set_ylabel('Accuracy %')\nax1.set_xlabel('Epoch')\nax1.set_title('Accuracy')\nl1 = ax1.legend(loc=\"best\")\n\nax2.plot(epoch_list, history.history['loss'], label='Train Loss')\nax2.plot(epoch_list, history.history['val_loss'], label='Validation Loss')\nax2.set_xticks(np.arange(0, EPOCHS + 1, 5))\nax2.set_ylabel('Loss %')\nax2.set_xlabel('Epoch')\nax2.set_title('Loss')\nl2 = ax2.legend(loc=\"best\")\n\nax3.plot(epoch_list, history.history['accuracy'], label='Accuracy')\nax3.plot(epoch_list, history.history['precision'], label='Precision')\nax3.plot(epoch_list, history.history['recall'], label='Recall')\nax3.plot(epoch_list, history.history['f1_score'], label='F1 score')\nax3.plot(epoch_list, history.history['fbeta_score'], label='Fbeta score')\nax3.plot(epoch_list, history.history['fmeasure'], label='FMeasure')\nax3.set_xticks(np.arange(0, EPOCHS + 1, 5))\nax3.set_ylabel('Score')\nax3.set_xlabel('Epoch')\nax3.set_title('Performance')\nl3 = ax3.legend(loc=\"best\")\n\nax4.plot(epoch_list, kappa_score.val_kappas, label='Kappa score')\nax4.set_xticks(np.arange(0, EPOCHS + 1, 5))\nax4.set_ylabel('Score')\nax4.set_xlabel('Epoch')\nax4.set_title('Kappa Metrics')\nl4 = ax4.legend(loc=\"best\")\n\ndisplay(\"Maximum Kappa Score: %s\" %max(kappa_score.val_kappas))","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T07:34:30.538741Z","iopub.execute_input":"2025-01-14T07:34:30.538969Z","iopub.status.idle":"2025-01-14T07:34:31.615407Z","shell.execute_reply.started":"2025-01-14T07:34:30.53895Z","shell.execute_reply":"2025-01-14T07:34:31.614605Z"}},"outputs":[{"output_type":"display_data","data":{"text/plain":"'Maximum Kappa Score: 0.902656107897868'"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}}],"execution_count":29},{"cell_type":"code","source":"import tensorflow as tf\nimport seaborn as sns\nfrom sklearn.metrics import confusion_matrix\nbest_model = tf.keras.models.load_model('/kaggle/working/model.h5', custom_objects={'precision': precision, 'recall': recall, 'fbeta_score': fbeta_score, 'fmeasure': fmeasure, 'mean_pred': mean_pred, 'f1_score': f1_score})\n\ny_pred_prob = best_model.predict(x_spval)\ny_pred = (y_pred_prob > 0.5).astype(int)\ny_true = y_spval\n","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T07:34:31.616369Z","iopub.execute_input":"2025-01-14T07:34:31.616756Z","iopub.status.idle":"2025-01-14T07:34:58.937919Z","shell.execute_reply.started":"2025-01-14T07:34:31.616718Z","shell.execute_reply":"2025-01-14T07:34:58.937161Z"}},"outputs":[{"name":"stdout","text":"\u001b[1m85/85\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m23s\u001b[0m 184ms/step\n","output_type":"stream"}],"execution_count":30},{"cell_type":"code","source":"from sklearn.metrics import roc_curve, auc\nfpr = dict()\ntpr = dict()\nroc_auc = dict()\nfor i in range(5):\n fpr[i], tpr[i], _ = roc_curve(y_true[:, i], y_pred_prob[:, i])\n roc_auc[i] = auc(fpr[i], tpr[i])\n\nplt.figure(figsize=(8, 6))\nfor i in range(5):\n plt.plot(fpr[i], tpr[i], label=f'ROC curve (area = {roc_auc[i]:.2f}) for class {i}')\nplt.plot([0, 1], [0, 1], 'k--')\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)')\nplt.legend(loc=\"lower right\")\nplt.show()","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T07:34:58.938688Z","iopub.execute_input":"2025-01-14T07:34:58.939218Z","iopub.status.idle":"2025-01-14T07:34:59.227063Z","shell.execute_reply.started":"2025-01-14T07:34:58.939192Z","shell.execute_reply":"2025-01-14T07:34:59.226243Z"}},"outputs":[{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}}],"execution_count":31},{"cell_type":"code","source":"y_true_flat = y_true.flatten()\ny_pred_flat = y_pred.flatten()\n\naccuracy = accuracy_score(y_true_flat, y_pred_flat)\nprint(f\"Accuracy: {accuracy:.4f}\")\n\nfor i in range(5):\n true_positives = np.sum((y_true[:, i] == 1) & (y_pred[:, i] == 1))\n false_negatives = np.sum((y_true[:, i] == 1) & (y_pred[:, i] == 0))\n true_negatives = np.sum((y_true[:, i] == 0) & (y_pred[:, i] == 0))\n false_positives = np.sum((y_true[:, i] == 0) & (y_pred[:, i] == 1))\n\n sensitivity = true_positives / (true_positives + false_negatives) if (true_positives + false_negatives) > 0 else 0\n specificity = true_negatives / (true_negatives + false_positives) if (true_negatives + false_positives) > 0 else 0\n\n print(f\"Class {i}:\")\n print(f\" Sensitivity: {sensitivity:.4f}\")\n print(f\" Specificity: {specificity:.4f}\")","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T07:34:59.227891Z","iopub.execute_input":"2025-01-14T07:34:59.228194Z","iopub.status.idle":"2025-01-14T07:34:59.24033Z","shell.execute_reply.started":"2025-01-14T07:34:59.228163Z","shell.execute_reply":"2025-01-14T07:34:59.239425Z"}},"outputs":[{"name":"stdout","text":"Accuracy: 0.9498\nClass 0:\n Sensitivity: 1.0000\n Specificity: 0.0000\nClass 1:\n Sensitivity: 0.9930\n Specificity: 0.9873\nClass 2:\n Sensitivity: 0.9137\n Specificity: 0.9305\nClass 3:\n Sensitivity: 0.8719\n Specificity: 0.9480\nClass 4:\n Sensitivity: 0.7678\n Specificity: 0.9538\n","output_type":"stream"}],"execution_count":32},{"cell_type":"code","source":"from tensorflow.keras.models import Model\ny_test = model.predict(x_test) > 0.5\ny_test = y_test.astype(int).sum(axis=1) - 1\n\ntest_df['diagnosis'] = y_test\ntest_df.to_csv('/kaggle/working/submission.csv',index=False)\ndisplay(test_df.head(5))\n\nimport datetime\nprint(\"Ran at UTC : \", datetime.datetime.utcnow())","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2025-01-14T07:34:59.241211Z","iopub.execute_input":"2025-01-14T07:34:59.241459Z","iopub.status.idle":"2025-01-14T07:35:16.257878Z","shell.execute_reply.started":"2025-01-14T07:34:59.241425Z","shell.execute_reply":"2025-01-14T07:35:16.256991Z"}},"outputs":[{"name":"stdout","text":"\u001b[1m61/61\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m16s\u001b[0m 269ms/step\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":" id_code diagnosis\n0 0005cfc8afb6 1\n1 003f0afdcd15 3\n2 006efc72b638 2\n3 00836aaacf06 2\n4 009245722fa4 2","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
id_codediagnosis
00005cfc8afb61
1003f0afdcd153
2006efc72b6382
300836aaacf062
4009245722fa42
\n
"},"metadata":{}},{"name":"stdout","text":"Ran at UTC : 2025-01-14 07:35:16.254169\n","output_type":"stream"}],"execution_count":33}]}