diff --git "a/Efficient-AD.ipynb" "b/Efficient-AD.ipynb"
new file mode 100644--- /dev/null
+++ "b/Efficient-AD.ipynb"
@@ -0,0 +1 @@
+{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.13","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"gpu","dataSources":[{"sourceId":3364939,"sourceType":"datasetVersion","datasetId":2029496}],"dockerImageVersionId":30648,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import warnings\nwarnings.filterwarnings(\"ignore\")","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:07:27.071603Z","iopub.execute_input":"2024-02-05T10:07:27.071953Z","iopub.status.idle":"2024-02-05T10:07:27.083430Z","shell.execute_reply.started":"2024-02-05T10:07:27.071923Z","shell.execute_reply":"2024-02-05T10:07:27.082646Z"},"trusted":true},"execution_count":1,"outputs":[]},{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\nimport tensorflow as tf\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nfrom sklearn.metrics import confusion_matrix, classification_report","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:07:27.085274Z","iopub.execute_input":"2024-02-05T10:07:27.085927Z","iopub.status.idle":"2024-02-05T10:07:42.023602Z","shell.execute_reply.started":"2024-02-05T10:07:27.085894Z","shell.execute_reply":"2024-02-05T10:07:42.022797Z"},"trusted":true},"execution_count":2,"outputs":[{"name":"stderr","text":"2024-02-05 10:07:29.737769: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered\n2024-02-05 10:07:29.737873: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered\n2024-02-05 10:07:29.892781: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n","output_type":"stream"}]},{"cell_type":"code","source":"PATH = \"/kaggle/input/alzheimer-mri-dataset/Dataset/\"\n\nBATCH_SIZE = 32\nIMG_SIZE = 224\n\nTRAIN_DS = tf.keras.utils.image_dataset_from_directory(\n  PATH,\n  validation_split=0.2,\n  subset=\"training\",\n  seed=123,\n  image_size=(IMG_SIZE, IMG_SIZE),\n  batch_size=BATCH_SIZE)\n\nVALID_DS = tf.keras.utils.image_dataset_from_directory(\n  PATH,\n  validation_split=0.2,\n  subset=\"validation\",\n  seed=123,\n  image_size=(IMG_SIZE, IMG_SIZE),\n  batch_size=BATCH_SIZE)\n\nCLASSES = TRAIN_DS.class_names\n\ncount = np.zeros(len(CLASSES), dtype=np.int32)\nfor _, labels in TRAIN_DS:\n    y, _, c = tf.unique_with_counts(labels)\n    count[y.numpy()] += c.numpy()    \nclass_weight = dict(enumerate(count))\n\nAUTOTUNE = tf.data.AUTOTUNE\n\nTRAIN_DS = TRAIN_DS.cache().prefetch(buffer_size=AUTOTUNE)\nVALID_DS = VALID_DS.cache().prefetch(buffer_size=AUTOTUNE)","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2024-02-05T10:07:42.025052Z","iopub.execute_input":"2024-02-05T10:07:42.025526Z","iopub.status.idle":"2024-02-05T10:08:15.980444Z","shell.execute_reply.started":"2024-02-05T10:07:42.025500Z","shell.execute_reply":"2024-02-05T10:08:15.979302Z"},"trusted":true},"execution_count":3,"outputs":[{"name":"stdout","text":"Found 6400 files belonging to 4 classes.\nUsing 5120 files for training.\nFound 6400 files belonging to 4 classes.\nUsing 1280 files for validation.\n","output_type":"stream"}]},{"cell_type":"code","source":"model = tf.keras.models.Sequential(name=\"DeepReLUwithEfficientNetB0\")\nmodel.add(tf.keras.applications.efficientnet.EfficientNetB0(include_top=False, weights='imagenet', input_shape=(224, 224, 3)))\nmodel.add(tf.keras.layers.Flatten())\nmodel.add(tf.keras.layers.Dense(4096, activation='relu'))\nmodel.add(tf.keras.layers.Dense(1024, activation='relu'))\nmodel.add(tf.keras.layers.Dense(256, activation='relu'))\nmodel.add(tf.keras.layers.Dense(64, activation='relu'))\nmodel.add(tf.keras.layers.Dense(len(CLASSES), activation='softmax'))\nmodel.layers[0].trainable = False\nmodel.summary()","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:08:15.981677Z","iopub.execute_input":"2024-02-05T10:08:15.981979Z","iopub.status.idle":"2024-02-05T10:08:20.540408Z","shell.execute_reply.started":"2024-02-05T10:08:15.981956Z","shell.execute_reply":"2024-02-05T10:08:20.539518Z"},"trusted":true},"execution_count":4,"outputs":[{"name":"stdout","text":"Downloading data from https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5\n16705208/16705208 [==============================] - 1s 0us/step\nModel: \"DeepReLUwithEfficientNetB0\"\n_________________________________________________________________\n Layer (type)                Output Shape              Param #   \n=================================================================\n efficientnetb0 (Functional  (None, 7, 7, 1280)        4049571   \n )                                                               \n                                                                 \n flatten (Flatten)           (None, 62720)             0         \n                                                                 \n dense (Dense)               (None, 4096)              256905216 \n                                                                 \n dense_1 (Dense)             (None, 1024)              4195328   \n                                                                 \n dense_2 (Dense)             (None, 256)               262400    \n                                                                 \n dense_3 (Dense)             (None, 64)                16448     \n                                                                 \n dense_4 (Dense)             (None, 4)                 260       \n                                                                 \n=================================================================\nTotal params: 265429223 (1012.53 MB)\nTrainable params: 261379652 (997.08 MB)\nNon-trainable params: 4049571 (15.45 MB)\n_________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"code","source":"model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.0001), loss=tf.keras.losses.SparseCategoricalCrossentropy(), metrics=[\"accuracy\"])\ntraining_history = model.fit(TRAIN_DS, validation_data=VALID_DS, epochs=30, class_weight=class_weight)\nresults = pd.DataFrame({\n    \"Training Accuracy\": training_history.history['accuracy'],\n    \"Validation Accuracy\": training_history.history['val_accuracy'],\n    \"Training Loss\": training_history.history['loss'],\n    \"Validation Loss\": training_history.history['val_loss']\n})","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:08:20.542598Z","iopub.execute_input":"2024-02-05T10:08:20.542913Z","iopub.status.idle":"2024-02-05T10:15:17.251956Z","shell.execute_reply.started":"2024-02-05T10:08:20.542887Z","shell.execute_reply":"2024-02-05T10:15:17.250861Z"},"trusted":true},"execution_count":5,"outputs":[{"name":"stdout","text":"Epoch 1/30\n","output_type":"stream"},{"name":"stderr","text":"2024-02-05 10:08:27.242707: E tensorflow/core/grappler/optimizers/meta_optimizer.cc:961] layout failed: INVALID_ARGUMENT: Size of values 0 does not match size of permutation 4 @ fanin shape inDeepReLUwithEfficientNetB0/efficientnetb0/block2b_drop/dropout/SelectV2-2-TransposeNHWCToNCHW-LayoutOptimizer\nWARNING: All log messages before absl::InitializeLog() is called are written to STDERR\nI0000 00:00:1707127710.375584      98 device_compiler.h:186] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.\n","output_type":"stream"},{"name":"stdout","text":"160/160 [==============================] - 28s 108ms/step - loss: 1747.6926 - accuracy: 0.5611 - val_loss: 1.0625 - val_accuracy: 0.5992\nEpoch 2/30\n160/160 [==============================] - 13s 83ms/step - loss: 1213.0515 - accuracy: 0.6561 - val_loss: 0.9330 - val_accuracy: 0.6586\nEpoch 3/30\n160/160 [==============================] - 13s 83ms/step - loss: 1016.2256 - accuracy: 0.7178 - val_loss: 0.7509 - val_accuracy: 0.6969\nEpoch 4/30\n160/160 [==============================] - 13s 83ms/step - loss: 836.8379 - accuracy: 0.7652 - val_loss: 0.6783 - val_accuracy: 0.7305\nEpoch 5/30\n160/160 [==============================] - 13s 84ms/step - loss: 740.5279 - accuracy: 0.7930 - val_loss: 0.5992 - val_accuracy: 0.7609\nEpoch 6/30\n160/160 [==============================] - 14s 85ms/step - loss: 625.0701 - accuracy: 0.8283 - val_loss: 0.5973 - val_accuracy: 0.7750\nEpoch 7/30\n160/160 [==============================] - 14s 85ms/step - loss: 605.4460 - accuracy: 0.8400 - val_loss: 0.7024 - val_accuracy: 0.7336\nEpoch 8/30\n160/160 [==============================] - 13s 84ms/step - loss: 542.1423 - accuracy: 0.8494 - val_loss: 0.7856 - val_accuracy: 0.7250\nEpoch 9/30\n160/160 [==============================] - 13s 84ms/step - loss: 490.5836 - accuracy: 0.8740 - val_loss: 0.4441 - val_accuracy: 0.8344\nEpoch 10/30\n160/160 [==============================] - 13s 84ms/step - loss: 591.7508 - accuracy: 0.8471 - val_loss: 0.4557 - val_accuracy: 0.8227\nEpoch 11/30\n160/160 [==============================] - 13s 84ms/step - loss: 512.5284 - accuracy: 0.8678 - val_loss: 0.4983 - val_accuracy: 0.8188\nEpoch 12/30\n160/160 [==============================] - 13s 84ms/step - loss: 543.9979 - accuracy: 0.8611 - val_loss: 0.3777 - val_accuracy: 0.8641\nEpoch 13/30\n160/160 [==============================] - 13s 84ms/step - loss: 404.7918 - accuracy: 0.8980 - val_loss: 0.6011 - val_accuracy: 0.8156\nEpoch 14/30\n160/160 [==============================] - 13s 84ms/step - loss: 328.3577 - accuracy: 0.9158 - val_loss: 0.5853 - val_accuracy: 0.8219\nEpoch 15/30\n160/160 [==============================] - 13s 84ms/step - loss: 337.4089 - accuracy: 0.9061 - val_loss: 0.3066 - val_accuracy: 0.8922\nEpoch 16/30\n160/160 [==============================] - 13s 83ms/step - loss: 353.0742 - accuracy: 0.8992 - val_loss: 0.5828 - val_accuracy: 0.8102\nEpoch 17/30\n160/160 [==============================] - 13s 83ms/step - loss: 373.5591 - accuracy: 0.9031 - val_loss: 0.4683 - val_accuracy: 0.8414\nEpoch 18/30\n160/160 [==============================] - 13s 83ms/step - loss: 407.0420 - accuracy: 0.8854 - val_loss: 0.3643 - val_accuracy: 0.8664\nEpoch 19/30\n160/160 [==============================] - 13s 83ms/step - loss: 315.1767 - accuracy: 0.9207 - val_loss: 0.2875 - val_accuracy: 0.8969\nEpoch 20/30\n160/160 [==============================] - 13s 83ms/step - loss: 286.8930 - accuracy: 0.9207 - val_loss: 0.3044 - val_accuracy: 0.8922\nEpoch 21/30\n160/160 [==============================] - 13s 83ms/step - loss: 298.9734 - accuracy: 0.9213 - val_loss: 0.3577 - val_accuracy: 0.8711\nEpoch 22/30\n160/160 [==============================] - 13s 84ms/step - loss: 227.9515 - accuracy: 0.9393 - val_loss: 0.8904 - val_accuracy: 0.7680\nEpoch 23/30\n160/160 [==============================] - 13s 83ms/step - loss: 274.1800 - accuracy: 0.9254 - val_loss: 0.5838 - val_accuracy: 0.7859\nEpoch 24/30\n160/160 [==============================] - 13s 84ms/step - loss: 217.6690 - accuracy: 0.9369 - val_loss: 0.8173 - val_accuracy: 0.7523\nEpoch 25/30\n160/160 [==============================] - 13s 84ms/step - loss: 206.3597 - accuracy: 0.9432 - val_loss: 0.9071 - val_accuracy: 0.7437\nEpoch 26/30\n160/160 [==============================] - 13s 84ms/step - loss: 170.0473 - accuracy: 0.9520 - val_loss: 0.1931 - val_accuracy: 0.9414\nEpoch 27/30\n160/160 [==============================] - 13s 84ms/step - loss: 164.3756 - accuracy: 0.9490 - val_loss: 0.2691 - val_accuracy: 0.9094\nEpoch 28/30\n160/160 [==============================] - 13s 84ms/step - loss: 178.7123 - accuracy: 0.9455 - val_loss: 0.2502 - val_accuracy: 0.9266\nEpoch 29/30\n160/160 [==============================] - 13s 84ms/step - loss: 196.2856 - accuracy: 0.9445 - val_loss: 0.2978 - val_accuracy: 0.9039\nEpoch 30/30\n160/160 [==============================] - 13s 83ms/step - loss: 173.2417 - accuracy: 0.9539 - val_loss: 0.2176 - val_accuracy: 0.9375\n","output_type":"stream"}]},{"cell_type":"code","source":"model.layers[0].trainable = True\nmodel.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.0001), loss=tf.keras.losses.SparseCategoricalCrossentropy(), metrics=[\"accuracy\"])\ntraining_history = model.fit(TRAIN_DS, validation_data=VALID_DS, epochs=15, class_weight=class_weight)\nresults_unfrozen = pd.DataFrame({\n    \"Training Accuracy\": training_history.history['accuracy'],\n    \"Validation Accuracy\": training_history.history['val_accuracy'],\n    \"Training Loss\": training_history.history['loss'],\n    \"Validation Loss\": training_history.history['val_loss']\n})","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:15:17.253510Z","iopub.execute_input":"2024-02-05T10:15:17.253828Z","iopub.status.idle":"2024-02-05T10:25:47.717153Z","shell.execute_reply.started":"2024-02-05T10:15:17.253801Z","shell.execute_reply":"2024-02-05T10:25:47.716228Z"},"trusted":true},"execution_count":6,"outputs":[{"name":"stdout","text":"Epoch 1/15\n","output_type":"stream"},{"name":"stderr","text":"2024-02-05 10:15:34.613508: E tensorflow/core/grappler/optimizers/meta_optimizer.cc:961] layout failed: INVALID_ARGUMENT: Size of values 0 does not match size of permutation 4 @ fanin shape inDeepReLUwithEfficientNetB0/efficientnetb0/block2b_drop/dropout/SelectV2-2-TransposeNHWCToNCHW-LayoutOptimizer\n","output_type":"stream"},{"name":"stdout","text":"160/160 [==============================] - 78s 257ms/step - loss: 1296.4480 - accuracy: 0.6305 - val_loss: 0.8041 - val_accuracy: 0.6453\nEpoch 2/15\n160/160 [==============================] - 39s 246ms/step - loss: 530.1741 - accuracy: 0.8516 - val_loss: 0.9093 - val_accuracy: 0.6656\nEpoch 3/15\n160/160 [==============================] - 39s 246ms/step - loss: 285.2955 - accuracy: 0.9262 - val_loss: 0.9554 - val_accuracy: 0.7109\nEpoch 4/15\n160/160 [==============================] - 39s 247ms/step - loss: 140.8667 - accuracy: 0.9605 - val_loss: 0.6218 - val_accuracy: 0.8211\nEpoch 5/15\n160/160 [==============================] - 39s 246ms/step - loss: 115.8797 - accuracy: 0.9627 - val_loss: 0.6105 - val_accuracy: 0.8391\nEpoch 6/15\n160/160 [==============================] - 39s 247ms/step - loss: 89.4116 - accuracy: 0.9723 - val_loss: 0.3307 - val_accuracy: 0.8992\nEpoch 7/15\n160/160 [==============================] - 39s 244ms/step - loss: 46.2791 - accuracy: 0.9801 - val_loss: 0.5520 - val_accuracy: 0.8602\nEpoch 8/15\n160/160 [==============================] - 39s 244ms/step - loss: 74.8651 - accuracy: 0.9729 - val_loss: 0.2972 - val_accuracy: 0.9219\nEpoch 9/15\n160/160 [==============================] - 39s 246ms/step - loss: 51.3475 - accuracy: 0.9787 - val_loss: 0.3490 - val_accuracy: 0.9172\nEpoch 10/15\n160/160 [==============================] - 39s 246ms/step - loss: 42.0641 - accuracy: 0.9809 - val_loss: 0.2000 - val_accuracy: 0.9320\nEpoch 11/15\n160/160 [==============================] - 40s 247ms/step - loss: 45.2430 - accuracy: 0.9834 - val_loss: 0.3072 - val_accuracy: 0.9180\nEpoch 12/15\n160/160 [==============================] - 40s 248ms/step - loss: 43.4032 - accuracy: 0.9854 - val_loss: 0.2312 - val_accuracy: 0.9203\nEpoch 13/15\n160/160 [==============================] - 40s 248ms/step - loss: 45.3657 - accuracy: 0.9855 - val_loss: 0.7804 - val_accuracy: 0.8469\nEpoch 14/15\n160/160 [==============================] - 40s 248ms/step - loss: 61.7115 - accuracy: 0.9832 - val_loss: 0.1678 - val_accuracy: 0.9422\nEpoch 15/15\n160/160 [==============================] - 40s 248ms/step - loss: 36.5023 - accuracy: 0.9893 - val_loss: 0.2214 - val_accuracy: 0.9375\n","output_type":"stream"}]},{"cell_type":"code","source":"model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.00001), loss=tf.keras.losses.SparseCategoricalCrossentropy(), metrics=[\"accuracy\"])\ntraining_history = model.fit(TRAIN_DS, validation_data=VALID_DS, epochs=15, class_weight=class_weight)\nresults_unfrozen_low_lr = pd.DataFrame({\n    \"Training Accuracy\": training_history.history['accuracy'],\n    \"Validation Accuracy\": training_history.history['val_accuracy'],\n    \"Training Loss\": training_history.history['loss'],\n    \"Validation Loss\": training_history.history['val_loss']\n})","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:25:47.718447Z","iopub.execute_input":"2024-02-05T10:25:47.718833Z","iopub.status.idle":"2024-02-05T10:36:21.355179Z","shell.execute_reply.started":"2024-02-05T10:25:47.718803Z","shell.execute_reply":"2024-02-05T10:36:21.354223Z"},"trusted":true},"execution_count":7,"outputs":[{"name":"stdout","text":"Epoch 1/15\n","output_type":"stream"},{"name":"stderr","text":"2024-02-05 10:26:04.880232: E tensorflow/core/grappler/optimizers/meta_optimizer.cc:961] layout failed: INVALID_ARGUMENT: Size of values 0 does not match size of permutation 4 @ fanin shape inDeepReLUwithEfficientNetB0/efficientnetb0/block2b_drop/dropout/SelectV2-2-TransposeNHWCToNCHW-LayoutOptimizer\n","output_type":"stream"},{"name":"stdout","text":"160/160 [==============================] - 74s 257ms/step - loss: 15.5505 - accuracy: 0.9967 - val_loss: 0.0863 - val_accuracy: 0.9734\nEpoch 2/15\n160/160 [==============================] - 40s 250ms/step - loss: 10.9042 - accuracy: 0.9979 - val_loss: 0.0879 - val_accuracy: 0.9711\nEpoch 3/15\n160/160 [==============================] - 40s 250ms/step - loss: 2.1416 - accuracy: 0.9992 - val_loss: 0.0808 - val_accuracy: 0.9758\nEpoch 4/15\n160/160 [==============================] - 40s 248ms/step - loss: 3.6791 - accuracy: 0.9992 - val_loss: 0.1035 - val_accuracy: 0.9758\nEpoch 5/15\n160/160 [==============================] - 40s 250ms/step - loss: 2.8339 - accuracy: 0.9994 - val_loss: 0.0856 - val_accuracy: 0.9742\nEpoch 6/15\n160/160 [==============================] - 40s 250ms/step - loss: 6.7326 - accuracy: 0.9984 - val_loss: 0.0852 - val_accuracy: 0.9828\nEpoch 7/15\n160/160 [==============================] - 40s 250ms/step - loss: 5.5398 - accuracy: 0.9988 - val_loss: 0.1281 - val_accuracy: 0.9672\nEpoch 8/15\n160/160 [==============================] - 40s 250ms/step - loss: 4.8523 - accuracy: 0.9986 - val_loss: 0.1352 - val_accuracy: 0.9680\nEpoch 9/15\n160/160 [==============================] - 40s 250ms/step - loss: 5.3067 - accuracy: 0.9986 - val_loss: 0.1177 - val_accuracy: 0.9727\nEpoch 10/15\n160/160 [==============================] - 40s 250ms/step - loss: 0.6918 - accuracy: 0.9994 - val_loss: 0.0887 - val_accuracy: 0.9805\nEpoch 11/15\n160/160 [==============================] - 40s 249ms/step - loss: 6.5639 - accuracy: 0.9994 - val_loss: 0.1224 - val_accuracy: 0.9727\nEpoch 12/15\n160/160 [==============================] - 40s 250ms/step - loss: 7.6069 - accuracy: 0.9986 - val_loss: 0.0874 - val_accuracy: 0.9766\nEpoch 13/15\n160/160 [==============================] - 40s 250ms/step - loss: 4.3595 - accuracy: 0.9990 - val_loss: 0.0670 - val_accuracy: 0.9828\nEpoch 14/15\n160/160 [==============================] - 40s 249ms/step - loss: 3.0697 - accuracy: 0.9990 - val_loss: 0.0677 - val_accuracy: 0.9805\nEpoch 15/15\n160/160 [==============================] - 40s 249ms/step - loss: 2.0835 - accuracy: 0.9992 - val_loss: 0.0711 - val_accuracy: 0.9836\n","output_type":"stream"}]},{"cell_type":"code","source":"training_history = model.fit(TRAIN_DS, validation_data=VALID_DS, epochs=30, class_weight=class_weight,\n                             callbacks=[tf.keras.callbacks.EarlyStopping(monitor='loss', patience=3, restore_best_weights= True, baseline=0.99)])\n\nresults_unfrozen_low_lr_final = pd.DataFrame({\n    \"Training Accuracy\": training_history.history['accuracy'],\n    \"Validation Accuracy\": training_history.history['val_accuracy'],\n    \"Training Loss\": training_history.history['loss'],\n    \"Validation Loss\": training_history.history['val_loss']\n})","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:36:21.356476Z","iopub.execute_input":"2024-02-05T10:36:21.356858Z","iopub.status.idle":"2024-02-05T10:43:11.254215Z","shell.execute_reply.started":"2024-02-05T10:36:21.356808Z","shell.execute_reply":"2024-02-05T10:43:11.253411Z"},"trusted":true},"execution_count":8,"outputs":[{"name":"stdout","text":"Epoch 1/15\n160/160 [==============================] - 43s 267ms/step - loss: 5.7842 - accuracy: 0.9992 - val_loss: 0.1263 - val_accuracy: 0.9711\nEpoch 2/15\n160/160 [==============================] - 41s 259ms/step - loss: 3.6281 - accuracy: 0.9990 - val_loss: 0.0970 - val_accuracy: 0.9805\nEpoch 3/15\n160/160 [==============================] - 41s 258ms/step - loss: 3.5063 - accuracy: 0.9992 - val_loss: 0.0676 - val_accuracy: 0.9836\nEpoch 4/15\n160/160 [==============================] - 41s 258ms/step - loss: 1.6296 - accuracy: 0.9996 - val_loss: 0.0885 - val_accuracy: 0.9797\nEpoch 5/15\n160/160 [==============================] - 40s 249ms/step - loss: 2.7224 - accuracy: 0.9988 - val_loss: 0.0515 - val_accuracy: 0.9883\nEpoch 6/15\n160/160 [==============================] - 41s 257ms/step - loss: 1.0344 - accuracy: 0.9994 - val_loss: 0.0670 - val_accuracy: 0.9891\nEpoch 7/15\n160/160 [==============================] - 41s 258ms/step - loss: 0.4643 - accuracy: 0.9998 - val_loss: 0.0653 - val_accuracy: 0.9875\nEpoch 8/15\n160/160 [==============================] - 40s 248ms/step - loss: 1.4863 - accuracy: 0.9994 - val_loss: 0.0630 - val_accuracy: 0.9883\nEpoch 9/15\n160/160 [==============================] - 40s 250ms/step - loss: 2.5791 - accuracy: 0.9994 - val_loss: 0.0535 - val_accuracy: 0.9906\nEpoch 10/15\n160/160 [==============================] - 42s 260ms/step - loss: 0.5794 - accuracy: 0.9998 - val_loss: 0.0709 - val_accuracy: 0.9867\n","output_type":"stream"}]},{"cell_type":"code","source":"import numpy as np\npredictions = []\nlabels =  []\nfor x, y in VALID_DS:\n    predictions.append(list(np.argmax(model.predict(x), axis = -1)))\n    labels.append(list(y.numpy()))\n\npredictions_final = []\ntrue_labels = []\nfor l in predictions:\n    predictions_final += l\nfor l in labels:\n    true_labels += l","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:43:11.255324Z","iopub.execute_input":"2024-02-05T10:43:11.255636Z","iopub.status.idle":"2024-02-05T10:43:16.954129Z","shell.execute_reply.started":"2024-02-05T10:43:11.255610Z","shell.execute_reply":"2024-02-05T10:43:16.953320Z"},"trusted":true},"execution_count":9,"outputs":[{"name":"stdout","text":"1/1 [==============================] - 1s 1s/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 43ms/step\n1/1 [==============================] - 0s 42ms/step\n1/1 [==============================] - 0s 43ms/step\n1/1 [==============================] - 0s 44ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 44ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 32ms/step\n1/1 [==============================] - 0s 31ms/step\n1/1 [==============================] - 0s 41ms/step\n1/1 [==============================] - 0s 32ms/step\n1/1 [==============================] - 0s 44ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 42ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 34ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 40ms/step\n1/1 [==============================] - 0s 32ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 43ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 36ms/step\n1/1 [==============================] - 0s 35ms/step\n1/1 [==============================] - 0s 32ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 34ms/step\n1/1 [==============================] - 0s 36ms/step\n1/1 [==============================] - 0s 36ms/step\n1/1 [==============================] - 0s 33ms/step\n1/1 [==============================] - 0s 34ms/step\n1/1 [==============================] - 0s 34ms/step\n1/1 [==============================] - 0s 44ms/step\n1/1 [==============================] - 0s 33ms/step\n","output_type":"stream"}]},{"cell_type":"code","source":"print(CLASSES)\nCLASS_NAMES = ['Mild\\nDemented', 'Moderate\\nDemented', 'Non\\nDemented', 'Very Mild\\nDemented']","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:43:16.955245Z","iopub.execute_input":"2024-02-05T10:43:16.955522Z","iopub.status.idle":"2024-02-05T10:43:16.960290Z","shell.execute_reply.started":"2024-02-05T10:43:16.955498Z","shell.execute_reply":"2024-02-05T10:43:16.959349Z"},"trusted":true},"execution_count":10,"outputs":[{"name":"stdout","text":"['Mild_Demented', 'Moderate_Demented', 'Non_Demented', 'Very_Mild_Demented']\n","output_type":"stream"}]},{"cell_type":"code","source":"cm = confusion_matrix(true_labels, predictions_final)\nax = sns.heatmap(data=cm, annot=True, cmap='crest', fmt='g', xticklabels=CLASS_NAMES, yticklabels=CLASS_NAMES, linecolor='black', linewidth=0.5)\nfig = ax.get_figure()\nfig.savefig('confusion-matrix.pdf', bbox_inches='tight')","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:43:16.963028Z","iopub.execute_input":"2024-02-05T10:43:16.963346Z","iopub.status.idle":"2024-02-05T10:43:17.690562Z","shell.execute_reply.started":"2024-02-05T10:43:16.963311Z","shell.execute_reply":"2024-02-05T10:43:17.689648Z"},"trusted":true},"execution_count":11,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 2 Axes>","image/png":"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"},"metadata":{}}]},{"cell_type":"code","source":"print(classification_report(true_labels, predictions_final))","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:43:17.691896Z","iopub.execute_input":"2024-02-05T10:43:17.693520Z","iopub.status.idle":"2024-02-05T10:43:17.710390Z","shell.execute_reply.started":"2024-02-05T10:43:17.693482Z","shell.execute_reply":"2024-02-05T10:43:17.709554Z"},"trusted":true},"execution_count":12,"outputs":[{"name":"stdout","text":"              precision    recall  f1-score   support\n\n           0       0.98      0.99      0.99       170\n           1       1.00      1.00      1.00        13\n           2       0.99      0.99      0.99       654\n           3       0.98      0.98      0.98       443\n\n    accuracy                           0.99      1280\n   macro avg       0.99      0.99      0.99      1280\nweighted avg       0.99      0.99      0.99      1280\n\n","output_type":"stream"}]},{"cell_type":"code","source":"sns.set_style(\"white\")\nfinal_results = pd.concat([results, results_unfrozen, results_unfrozen_low_lr, results_unfrozen_low_lr_final], ignore_index=True).reset_index()\nfinal_results.columns = [\"Epochs\", \"Training Accuracy\", \"Validation Accuracy\", \"Training Loss\", \"Validation Loss\"]\nfinal_results[\"Epochs\"] += 1\nmelted_results = pd.melt(final_results, id_vars=['Epochs'], var_name='Metric', value_name='Value')\n\nfig = plt.figure(figsize=(6,4))\nfig.subplots_adjust(hspace=0.4, wspace=0.4)\n\nax = fig.add_subplot(2, 1, 1)\nax = sns.lineplot(data=melted_results[melted_results[\"Metric\"].isin([\"Training Accuracy\", \"Validation Accuracy\"])], y=\"Value\", x=\"Epochs\", hue=\"Metric\", linewidth = 2, ax=ax)\nax.set_xlabel(\"Epochs\", fontdict={'weight': 'bold'})\nax.set_ylabel(\"Accuracy\", fontdict={'weight': 'bold'})\nax.margins(x=0)\nsns.move_legend(ax, \"upper left\", bbox_to_anchor=(1, 1.06))\n\nax = fig.add_subplot(2, 1, 2)\nax = sns.lineplot(data=melted_results[melted_results[\"Metric\"].isin([\"Training Loss\", \"Validation Loss\"])], y=\"Value\", x=\"Epochs\", hue=\"Metric\", linewidth = 2, ax=ax)\nax.set_xlabel(\"Epochs\", fontdict={'weight': 'bold'})\nax.set_ylabel(\"Loss\", fontdict={'weight': 'bold'})\nax.margins(x=0)\nsns.move_legend(ax, \"upper left\", bbox_to_anchor=(1, 1.06))","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:43:17.711358Z","iopub.execute_input":"2024-02-05T10:43:17.711595Z","iopub.status.idle":"2024-02-05T10:43:18.377799Z","shell.execute_reply.started":"2024-02-05T10:43:17.711574Z","shell.execute_reply":"2024-02-05T10:43:18.376860Z"},"trusted":true},"execution_count":13,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 600x400 with 2 Axes>","image/png":"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"},"metadata":{}}]},{"cell_type":"code","source":"fig, ax = plt.subplots(figsize=(6,2.5))\nax = sns.lineplot(data=melted_results[melted_results[\"Metric\"].isin([\"Training Accuracy\", \"Validation Accuracy\"])], y=\"Value\", x=\"Epochs\", hue=\"Metric\", linewidth = 2, ax=ax)\nax.set_xlabel(\"Epochs\", fontdict={'weight': 'bold'})\nax.set_ylabel(\"Accuracy\", fontdict={'weight': 'bold'})\nax.margins(x=0)\nsns.move_legend(ax, \"upper left\", bbox_to_anchor=(1, 1.06))\nplt.vlines(x = 30, ymin = 0.5, ymax = 1.1, colors = 'purple', linestyles='dashed', linewidth=0.7)\nplt.text(10, 0.55, \"Frozen Base Model\", verticalalignment='center', fontdict={'fontsize': 'x-small', 'fontweight': 'bold'})\nplt.vlines(x = 45, ymin = 0.5, ymax = 1.1, colors = 'purple', linestyles='dashed', linewidth=0.7)\nplt.text(37.5, 0.5, \"Unfrozen\\nBase Model\", horizontalalignment='center', fontdict={'fontsize': 'x-small', 'fontweight': 'bold'})\nplt.text(55, 0.5, \"Reduced\\nLearning\\nRate\", horizontalalignment='center', fontdict={'fontsize': 'x-small', 'fontweight': 'bold'})\nplt.savefig('efficient-ad-training-accuracy.pdf', bbox_inches='tight')","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:43:18.379133Z","iopub.execute_input":"2024-02-05T10:43:18.379739Z","iopub.status.idle":"2024-02-05T10:43:19.029641Z","shell.execute_reply.started":"2024-02-05T10:43:18.379703Z","shell.execute_reply":"2024-02-05T10:43:19.028581Z"},"trusted":true},"execution_count":14,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 600x250 with 1 Axes>","image/png":"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"},"metadata":{}}]},{"cell_type":"code","source":"model.save(\"Efficient-AD.keras\")","metadata":{"execution":{"iopub.status.busy":"2024-02-05T10:43:19.030967Z","iopub.execute_input":"2024-02-05T10:43:19.031337Z","iopub.status.idle":"2024-02-05T10:43:32.900402Z","shell.execute_reply.started":"2024-02-05T10:43:19.031303Z","shell.execute_reply":"2024-02-05T10:43:32.899033Z"},"trusted":true},"execution_count":15,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}
\ No newline at end of file