brain-mri-multilevel-wavelet-cnn.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import os,shutil\nimport tensorflow \nfrom tensorflow import keras\nfrom tensorflow.keras import initializers, layers\nfrom tensorflow.keras.layers import *\nfrom matplotlib import pyplot as plt\nfrom tensorflow.keras import backend as K\nfrom tensorflow.keras import activations\nfrom tensorflow.keras.models import Model,Sequential\nfrom tensorflow.keras.optimizers import Adam\nfrom tensorflow.keras.optimizers import SGD\nfrom tensorflow.keras.optimizers import RMSprop\nfrom tensorflow.keras.applications.resnet50 import ResNet50, preprocess_input\nfrom tensorflow.keras.applications.vgg19 import VGG19\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator\nfrom tensorflow.keras.layers import Dense, GlobalAveragePooling2D\nfrom tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping\nfrom tensorflow.keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPool2D, BatchNormalization, AveragePooling2D, GlobalAveragePooling2D\nfrom tensorflow.keras.utils import to_categorical\nfrom tensorflow.keras.preprocessing.image import save_img\nfrom tensorflow.keras.preprocessing.image import load_img\nfrom tensorflow.keras.preprocessing.image import img_to_array\nfrom tensorflow.keras.preprocessing.image import array_to_img\nfrom tensorflow.keras import metrics\nfrom pathlib import Path\nimport sklearn.metrics\nimport numpy as np\nimport pandas as pd\nimport cv2\n\nimport seaborn as sns","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-02-06T09:48:44.408970Z","iopub.execute_input":"2022-02-06T09:48:44.409315Z","iopub.status.idle":"2022-02-06T09:48:50.248592Z","shell.execute_reply.started":"2022-02-06T09:48:44.409230Z","shell.execute_reply":"2022-02-06T09:48:50.247842Z"},"trusted":true},"execution_count":1,"outputs":[]},{"cell_type":"code","source":"#for reproducible results\nseed_value=42\nos.environ['PYTHONHASHSEED']=str(seed_value)\nimport random\nrandom.seed(seed_value)\nnp.random.seed(seed_value)\nimport tensorflow as tf\ntf.random.set_seed(seed_value)","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:48:50.251673Z","iopub.execute_input":"2022-02-06T09:48:50.252179Z","iopub.status.idle":"2022-02-06T09:48:50.257419Z","shell.execute_reply.started":"2022-02-06T09:48:50.252139Z","shell.execute_reply":"2022-02-06T09:48:50.256778Z"},"trusted":true},"execution_count":2,"outputs":[]},{"cell_type":"code","source":"#defining function for resizing and loading images\nIMAGE_SIZE = 224\ndef read_image(filepath):\n    return cv2.imread(filepath,0) \ndef resize_image(image, image_size):\n    return cv2.resize(image.copy(), image_size, interpolation=cv2.INTER_AREA)","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:48:50.260387Z","iopub.execute_input":"2022-02-06T09:48:50.260584Z","iopub.status.idle":"2022-02-06T09:48:50.267082Z","shell.execute_reply.started":"2022-02-06T09:48:50.260560Z","shell.execute_reply":"2022-02-06T09:48:50.266458Z"},"trusted":true},"execution_count":3,"outputs":[]},{"cell_type":"code","source":"def to_rgb(img):\n    \n    img_rgb = np.asarray(np.dstack((img, img, img)), dtype=np.uint8)\n    return img_rgb","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:48:50.269285Z","iopub.execute_input":"2022-02-06T09:48:50.269614Z","iopub.status.idle":"2022-02-06T09:48:50.276414Z","shell.execute_reply.started":"2022-02-06T09:48:50.269578Z","shell.execute_reply":"2022-02-06T09:48:50.275599Z"},"trusted":true},"execution_count":4,"outputs":[]},{"cell_type":"code","source":"def dataset():    \n    class_=['no','yes']\n    y=list()\n    x=list()\n    data=pd.DataFrame()\n    for i,j in enumerate(class_):\n        for n in os.listdir('../input/brain-tumor-detection/'+j):\n            y.append(i)\n            x.append('../input/brain-tumor-detection/'+j+'/'+n)\n\n    data=pd.DataFrame(list(zip(x,y)),columns=[\"image\",\"truth\"])     \n    return data\n","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:48:50.277630Z","iopub.execute_input":"2022-02-06T09:48:50.278322Z","iopub.status.idle":"2022-02-06T09:48:50.285576Z","shell.execute_reply.started":"2022-02-06T09:48:50.278267Z","shell.execute_reply":"2022-02-06T09:48:50.284792Z"},"trusted":true},"execution_count":5,"outputs":[]},{"cell_type":"code","source":"data=dataset()","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:48:50.286739Z","iopub.execute_input":"2022-02-06T09:48:50.287058Z","iopub.status.idle":"2022-02-06T09:48:50.606387Z","shell.execute_reply.started":"2022-02-06T09:48:50.287023Z","shell.execute_reply":"2022-02-06T09:48:50.605547Z"},"trusted":true},"execution_count":6,"outputs":[]},{"cell_type":"code","source":"#wavelet transform\nimport pywt\ndef w2d(img, mode='haar', level=1):\n    imArray = img\n    #Datatype conversions\n    #convert to grayscale\n   # imArray = cv2.cvtColor( imArray,cv2.COLOR_RGB2GRAY )\n    #convert to float\n    imArray =  np.float32(imArray)   \n    imArray /= 255;\n    # compute coefficients \n    coeffs=pywt.wavedec2(imArray, mode, level=level)\n\n    #Process Coefficients\n    coeffs_H=list(coeffs)  \n    coeffs_H[0] *= 0;  \n\n    # reconstruction\n    imArray_H=pywt.waverec2(coeffs_H, mode);\n    imArray_H *= 255;\n    imArray_H =  np.uint8(imArray_H)\n\n    return imArray_H","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:48:50.607636Z","iopub.execute_input":"2022-02-06T09:48:50.607962Z","iopub.status.idle":"2022-02-06T09:48:50.662381Z","shell.execute_reply.started":"2022-02-06T09:48:50.607924Z","shell.execute_reply":"2022-02-06T09:48:50.661683Z"},"trusted":true},"execution_count":7,"outputs":[]},{"cell_type":"code","source":"def preprocess(data) :   \n    X=np.zeros((data.shape[0],IMAGE_SIZE,IMAGE_SIZE))\n    rgb_list=[]\n    Y=[]\n    for i in range(data.shape[0]):\n        image=read_image(data['image'][i])\n        if image is not None:\n\n            X[i]=resize_image(image,(IMAGE_SIZE,IMAGE_SIZE))\n            #X[i]=w2d(X[i],'db1',3)\n            rgb = to_rgb(X[i])\n\n            rgb_list.append(rgb/255.0)\n\n            Y.append(data['truth'][i])\n\n    rgb_arr = np.stack([rgb_list],axis=4)\n    rgb_arr_to_3d = np.squeeze(rgb_arr, axis=4)\n    X=rgb_arr_to_3d   \n    return X, Y\n        \n","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:48:50.663338Z","iopub.execute_input":"2022-02-06T09:48:50.665151Z","iopub.status.idle":"2022-02-06T09:48:50.672850Z","shell.execute_reply.started":"2022-02-06T09:48:50.665096Z","shell.execute_reply":"2022-02-06T09:48:50.672050Z"},"trusted":true},"execution_count":8,"outputs":[]},{"cell_type":"code","source":"X,Y= preprocess(data)","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:48:50.675705Z","iopub.execute_input":"2022-02-06T09:48:50.675954Z","iopub.status.idle":"2022-02-06T09:49:14.845263Z","shell.execute_reply.started":"2022-02-06T09:48:50.675918Z","shell.execute_reply":"2022-02-06T09:49:14.844468Z"},"trusted":true},"execution_count":9,"outputs":[]},{"cell_type":"code","source":"X.shape","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:14.848883Z","iopub.execute_input":"2022-02-06T09:49:14.849099Z","iopub.status.idle":"2022-02-06T09:49:14.855670Z","shell.execute_reply.started":"2022-02-06T09:49:14.849075Z","shell.execute_reply":"2022-02-06T09:49:14.855036Z"},"trusted":true},"execution_count":10,"outputs":[{"execution_count":10,"output_type":"execute_result","data":{"text/plain":"(3000, 224, 224, 3)"},"metadata":{}}]},{"cell_type":"code","source":"from sklearn.model_selection import train_test_split\nX_train, X_val, Y_train, Y_val = train_test_split(X, Y,stratify=Y, test_size=0.1,random_state=41)","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:14.856929Z","iopub.execute_input":"2022-02-06T09:49:14.857406Z","iopub.status.idle":"2022-02-06T09:49:15.914773Z","shell.execute_reply.started":"2022-02-06T09:49:14.857370Z","shell.execute_reply":"2022-02-06T09:49:15.914041Z"},"trusted":true},"execution_count":11,"outputs":[]},{"cell_type":"code","source":"Y_train= to_categorical(Y_train, num_classes=2)\nY_val= to_categorical(Y_val, num_classes=2)","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:15.916675Z","iopub.execute_input":"2022-02-06T09:49:15.917210Z","iopub.status.idle":"2022-02-06T09:49:15.922706Z","shell.execute_reply.started":"2022-02-06T09:49:15.917170Z","shell.execute_reply":"2022-02-06T09:49:15.922044Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"code","source":"from  matplotlib import pyplot as plt\nplt.imshow(X[0],cmap='gray')","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:15.923966Z","iopub.execute_input":"2022-02-06T09:49:15.924285Z","iopub.status.idle":"2022-02-06T09:49:16.190013Z","shell.execute_reply.started":"2022-02-06T09:49:15.924248Z","shell.execute_reply":"2022-02-06T09:49:16.189330Z"},"trusted":true},"execution_count":13,"outputs":[{"execution_count":13,"output_type":"execute_result","data":{"text/plain":"<matplotlib.image.AxesImage at 0x7f2ed1fc2650>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"###Capsule layer implementation#####3\n# the squashing function.\n\"\"\"\nThe non-linear activation used in Capsule. It drives the length of a large vector to near 1 and small vector to 0\n:param vectors: some vectors to be squashed, N-dim tensor\n:param axis: the axis to squash\n:return: a Tensor with same shape as input vectors\n\"\"\"\ndef squash(x, axis=-1):\n    s_squared_norm = K.sum(K.square(x), axis, keepdims=True) + K.epsilon()\n    scale = K.sqrt(s_squared_norm) / (0.5 + s_squared_norm)\n    return scale * x\n\n# define our own softmax function instead of K.softmax\n# because K.softmax can not specify axis.\ndef softmax(x, axis=-1):\n    ex = K.exp(x - K.max(x, axis=axis, keepdims=True))\n    return ex / K.sum(ex, axis=axis, keepdims=True)\ndef margin_loss(y_true, y_pred):\n    lamb, margin = 0.5, 0.1 #default lambda 0.5 - but test with lambda with 0.9 - 0.1\n    y_pred=K.flatten(y_pred)\n    y_true=K.flatten(y_true)\n    return K.mean(y_true * K.square(K.relu(1 - margin - y_pred)) + lamb * (1 - y_true) * K.square(K.relu(y_pred - margin)), axis=-1)\n\ndef caps_batch_dot(x, y):\n    x = K.expand_dims(x, 2)\n    if K.int_shape(x)[3] is not None:\n        y = K.permute_dimensions(y, (0, 1, 3, 2))\n    o = tensorflow.matmul(x, y)\n    return K.squeeze(o, 2)\n\nclass Capsule(Layer):\n    \"\"\"\n    The capsule layer. It is similar to Dense layer. Dense layer has `in_num` inputs, each is a scalar, the output of the \n    neuron from the former layer, and it has `out_num` output neurons. CapsuleLayer just expand the output of the neuron\n    from scalar to vector. So its input shape = [None, input_num_capsule, input_dim_capsule] and output shape = \\\n    [None, num_capsule, dim_capsule]. For Dense Layer, input_dim_capsule = dim_capsule = 1.\n    \n    :param num_capsule: number of capsules in this layer\n    :param dim_capsule: dimension of the output vectors of the capsules in this layer\n    :param routings: number of iterations for the routing algorithm\n    \"\"\"\n    def __init__(self,\n                 num_capsule,\n                 dim_capsule,\n                 routings=3,\n                 share_weights=True,\n                 activation='squash',\n                 **kwargs):\n        super(Capsule, self).__init__(**kwargs)\n        self.num_capsule = num_capsule\n        self.dim_capsule = dim_capsule\n        self.routings = routings\n        self.share_weights = share_weights\n        if activation == 'squash':\n            self.activation = squash\n        else:\n            self.activation = activations.get(activation)\n\n    def build(self, input_shape):\n        input_dim_capsule = input_shape[-1]\n        if self.share_weights:\n            self.kernel = self.add_weight(\n                name='capsule_kernel',\n                shape=(1, input_dim_capsule,\n                       self.num_capsule * self.dim_capsule),\n                initializer='glorot_uniform',\n                trainable=True)\n        else:\n            if input_shape[-2] is None:\n                raise ValueError(\"Input Shape must be defied if weights not shared.\")\n            input_num_capsule = input_shape[-2]\n            self.kernel = self.add_weight(\n                name='capsule_kernel',\n                shape=(input_num_capsule, input_dim_capsule,\n                       self.num_capsule * self.dim_capsule),\n                initializer='glorot_uniform',\n                trainable=True)\n\n    def call(self, inputs):\n        \"\"\"Following the routing algorithm from Hinton's paper,\n        but replace b = b + <u,v> with b = <u,v>.\n        This change can improve the feature representation of Capsule.\n        However, you can replace\n            b = K.batch_dot(outputs, hat_inputs, [2, 3])\n        with\n            b += K.batch_dot(outputs, hat_inputs, [2, 3])\n        to realize a standard routing.\n        \"\"\"\n\n        if self.share_weights:\n            hat_inputs = K.conv1d(inputs, self.kernel)\n        else:\n            hat_inputs = K.local_conv1d(inputs, self.kernel, [1], [1])\n\n        batch_size = K.shape(inputs)[0]\n        input_num_capsule = K.shape(inputs)[1]\n        hat_inputs = K.reshape(hat_inputs,\n                               (batch_size, input_num_capsule,\n                                self.num_capsule, self.dim_capsule))\n        hat_inputs = K.permute_dimensions(hat_inputs, (0, 2, 1, 3))\n\n        b = K.zeros_like(hat_inputs[:, :, :, 0])\n\n        \n        # Begin: Routing algorithm ---------------------------------------------------------------------#\n        # The prior for coupling coefficient, initialized as zeros.\n        # b.shape = [None, self.num_capsule, self.input_num_capsule].\n\n        for i in range(self.routings):\n            c = softmax(b, 1)\n            o = self.activation(caps_batch_dot(c, hat_inputs))\n            if i < self.routings - 1:\n                b = caps_batch_dot(o, hat_inputs)\n                if K.backend() == 'theano':\n                    o = K.sum(o, axis=1)\n        # End: Routing algorithm -----------------------------------------------------------------------#\n        return o\n\n    def compute_output_shape(self, input_shape):\n        return (None, self.num_capsule, self.dim_capsule)\n        \n    def get_config(self):\n        config = {\n            'num_capsule': self.num_capsule,\n            'dim_capsule': self.dim_capsule,\n            'routings': self.routings\n        }\n        base_config = super(Capsule, self).get_config()\n        return dict(list(base_config.items()) + list(config.items()))","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:16.191312Z","iopub.execute_input":"2022-02-06T09:49:16.191672Z","iopub.status.idle":"2022-02-06T09:49:16.214682Z","shell.execute_reply.started":"2022-02-06T09:49:16.191637Z","shell.execute_reply":"2022-02-06T09:49:16.214034Z"},"trusted":true},"execution_count":14,"outputs":[]},{"cell_type":"code","source":"# batch operation usng tensor slice\ndef WaveletTransformAxisY(batch_img):\n    odd_img  = batch_img[:,0::2]\n    even_img = batch_img[:,1::2]\n    L = (odd_img + even_img) / 2.0\n    H = K.abs(odd_img - even_img)\n    return L, H\n\ndef WaveletTransformAxisX(batch_img):\n    # transpose + fliplr\n    tmp_batch = K.permute_dimensions(batch_img, [0, 2, 1])[:,:,::-1]\n    _dst_L, _dst_H = WaveletTransformAxisY(tmp_batch)\n    # transpose + flipud\n    dst_L = K.permute_dimensions(_dst_L, [0, 2, 1])[:,::-1,...]\n    dst_H = K.permute_dimensions(_dst_H, [0, 2, 1])[:,::-1,...]\n    return dst_L, dst_H","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:16.217119Z","iopub.execute_input":"2022-02-06T09:49:16.217784Z","iopub.status.idle":"2022-02-06T09:49:16.228306Z","shell.execute_reply.started":"2022-02-06T09:49:16.217749Z","shell.execute_reply":"2022-02-06T09:49:16.227552Z"},"trusted":true},"execution_count":15,"outputs":[]},{"cell_type":"code","source":"def Wavelet(batch_image):\n    # make channel first image\n    batch_image = K.permute_dimensions(batch_image, [0, 3, 1, 2])\n    r = batch_image[:,0]\n    g = batch_image[:,1]\n    b = batch_image[:,2]\n\n    # level 1 decomposition\n    wavelet_L, wavelet_H = WaveletTransformAxisY(r)\n    r_wavelet_LL, r_wavelet_LH = WaveletTransformAxisX(wavelet_L)\n    r_wavelet_HL, r_wavelet_HH = WaveletTransformAxisX(wavelet_H)\n\n    wavelet_L, wavelet_H = WaveletTransformAxisY(g)\n    g_wavelet_LL, g_wavelet_LH = WaveletTransformAxisX(wavelet_L)\n    g_wavelet_HL, g_wavelet_HH = WaveletTransformAxisX(wavelet_H)\n\n    wavelet_L, wavelet_H = WaveletTransformAxisY(b)\n    b_wavelet_LL, b_wavelet_LH = WaveletTransformAxisX(wavelet_L)\n    b_wavelet_HL, b_wavelet_HH = WaveletTransformAxisX(wavelet_H)\n\n    wavelet_data = [r_wavelet_LL, r_wavelet_LH, r_wavelet_HL, r_wavelet_HH, \n                    g_wavelet_LL, g_wavelet_LH, g_wavelet_HL, g_wavelet_HH,\n                    b_wavelet_LL, b_wavelet_LH, b_wavelet_HL, b_wavelet_HH]\n    transform_batch = K.stack(wavelet_data, axis=1)\n\n    # level 2 decomposition\n    wavelet_L2, wavelet_H2 = WaveletTransformAxisY(r_wavelet_LL)\n    r_wavelet_LL2, r_wavelet_LH2 = WaveletTransformAxisX(wavelet_L2)\n    r_wavelet_HL2, r_wavelet_HH2 = WaveletTransformAxisX(wavelet_H2)\n\n    wavelet_L2, wavelet_H2 = WaveletTransformAxisY(g_wavelet_LL)\n    g_wavelet_LL2, g_wavelet_LH2 = WaveletTransformAxisX(wavelet_L2)\n    g_wavelet_HL2, g_wavelet_HH2 = WaveletTransformAxisX(wavelet_H2)\n\n    wavelet_L2, wavelet_H2 = WaveletTransformAxisY(b_wavelet_LL)\n    b_wavelet_LL2, b_wavelet_LH2 = WaveletTransformAxisX(wavelet_L2)\n    b_wavelet_HL2, b_wavelet_HH2 = WaveletTransformAxisX(wavelet_H2)\n\n\n    wavelet_data_l2 = [r_wavelet_LL2, r_wavelet_LH2, r_wavelet_HL2, r_wavelet_HH2, \n                    g_wavelet_LL2, g_wavelet_LH2, g_wavelet_HL2, g_wavelet_HH2,\n                    b_wavelet_LL2, b_wavelet_LH2, b_wavelet_HL2, b_wavelet_HH2]\n    transform_batch_l2 = K.stack(wavelet_data_l2, axis=1)\n\n    # level 3 decomposition\n    wavelet_L3, wavelet_H3 = WaveletTransformAxisY(r_wavelet_LL2)\n    r_wavelet_LL3, r_wavelet_LH3 = WaveletTransformAxisX(wavelet_L3)\n    r_wavelet_HL3, r_wavelet_HH3 = WaveletTransformAxisX(wavelet_H3)\n\n    wavelet_L3, wavelet_H3 = WaveletTransformAxisY(g_wavelet_LL2)\n    g_wavelet_LL3, g_wavelet_LH3 = WaveletTransformAxisX(wavelet_L3)\n    g_wavelet_HL3, g_wavelet_HH3 = WaveletTransformAxisX(wavelet_H3)\n\n    wavelet_L3, wavelet_H3 = WaveletTransformAxisY(b_wavelet_LL2)\n    b_wavelet_LL3, b_wavelet_LH3 = WaveletTransformAxisX(wavelet_L3)\n    b_wavelet_HL3, b_wavelet_HH3 = WaveletTransformAxisX(wavelet_H3)\n\n    wavelet_data_l3 = [r_wavelet_LL3, r_wavelet_LH3, r_wavelet_HL3, r_wavelet_HH3, \n                    g_wavelet_LL3, g_wavelet_LH3, g_wavelet_HL3, g_wavelet_HH3,\n                    b_wavelet_LL3, b_wavelet_LH3, b_wavelet_HL3, b_wavelet_HH3]\n    transform_batch_l3 = K.stack(wavelet_data_l3, axis=1)\n\n    # level 4 decomposition\n    wavelet_L4, wavelet_H4 = WaveletTransformAxisY(r_wavelet_LL3)\n    r_wavelet_LL4, r_wavelet_LH4 = WaveletTransformAxisX(wavelet_L4)\n    r_wavelet_HL4, r_wavelet_HH4 = WaveletTransformAxisX(wavelet_H4)\n\n    wavelet_L4, wavelet_H4 = WaveletTransformAxisY(g_wavelet_LL3)\n    g_wavelet_LL4, g_wavelet_LH4 = WaveletTransformAxisX(wavelet_L4)\n    g_wavelet_HL4, g_wavelet_HH4 = WaveletTransformAxisX(wavelet_H4)\n\n    wavelet_L3, wavelet_H3 = WaveletTransformAxisY(b_wavelet_LL3)\n    b_wavelet_LL4, b_wavelet_LH4 = WaveletTransformAxisX(wavelet_L4)\n    b_wavelet_HL4, b_wavelet_HH4 = WaveletTransformAxisX(wavelet_H4)\n\n\n    wavelet_data_l4 = [r_wavelet_LL4, r_wavelet_LH4, r_wavelet_HL4, r_wavelet_HH4, \n                    g_wavelet_LL4, g_wavelet_LH4, g_wavelet_HL4, g_wavelet_HH4,\n                    b_wavelet_LL4, b_wavelet_LH4, b_wavelet_HL4, b_wavelet_HH4]\n    transform_batch_l4 = K.stack(wavelet_data_l4, axis=1)\n\n    # print('shape before')\n    # print(transform_batch.shape)\n    # print(transform_batch_l2.shape)\n    # print(transform_batch_l3.shape)\n    # print(transform_batch_l4.shape)\n\n    decom_level_1 = K.permute_dimensions(transform_batch, [0, 2, 3, 1])\n    decom_level_2 = K.permute_dimensions(transform_batch_l2, [0, 2, 3, 1])\n    decom_level_3 = K.permute_dimensions(transform_batch_l3, [0, 2, 3, 1])\n    decom_level_4 = K.permute_dimensions(transform_batch_l4, [0, 2, 3, 1])\n    \n    # print('shape after')\n    # print(decom_level_1.shape)\n    # print(decom_level_2.shape)\n    # print(decom_level_3.shape)\n    # print(decom_level_4.shape)\n    return [decom_level_1, decom_level_2, decom_level_3, decom_level_4]\n\n\ndef Wavelet_out_shape(input_shapes):\n    # print('in to shape')\n    return [tuple([None, 112, 112, 12]), tuple([None, 56, 56, 12]), \n            tuple([None, 28, 28, 12]), tuple([None, 14, 14, 12])]","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:16.229723Z","iopub.execute_input":"2022-02-06T09:49:16.229980Z","iopub.status.idle":"2022-02-06T09:49:16.250674Z","shell.execute_reply.started":"2022-02-06T09:49:16.229946Z","shell.execute_reply":"2022-02-06T09:49:16.250039Z"},"trusted":true},"execution_count":16,"outputs":[]},{"cell_type":"code","source":"input_shape = 224, 224, 3\n\ninput_ = Input(input_shape, name='the_input')\n# wavelet = Lambda(Wavelet, name='wavelet')\nwavelet = Lambda(Wavelet, Wavelet_out_shape, name='wavelet')\ninput_l1, input_l2, input_l3, input_l4 = wavelet(input_)\n# print(input_l1)\n# print(input_l2)\n# print(input_l3)\n# print(input_l4)\n# level one decomposition starts\nconv_1 = Conv2D(64, kernel_size=(3, 3), padding='same', name='conv_1')(input_l1)\nnorm_1 = BatchNormalization(name='norm_1')(conv_1)\nrelu_1 = Activation('relu', name='relu_1')(norm_1)\n\nconv_1_2 = Conv2D(64, kernel_size=(3, 3), strides=(2, 2), padding='same', name='conv_1_2')(relu_1)\nnorm_1_2 = BatchNormalization(name='norm_1_2')(conv_1_2)\nrelu_1_2 = Activation('relu', name='relu_1_2')(norm_1_2)\n\n# level two decomposition starts\nconv_a = Conv2D(filters=64, kernel_size=(3, 3), padding='same', name='conv_a')(input_l2)\nnorm_a = BatchNormalization(name='norm_a')(conv_a)\nrelu_a = Activation('relu', name='relu_a')(norm_a)\n\n# concate level one and level two decomposition\nconcate_level_2 = concatenate([relu_1_2, relu_a])\nconv_2 = Conv2D(128, kernel_size=(3, 3), padding='same', name='conv_2')(concate_level_2)\nnorm_2 = BatchNormalization(name='norm_2')(conv_2)\nrelu_2 = Activation('relu', name='relu_2')(norm_2)\n\nconv_2_2 = Conv2D(128, kernel_size=(3, 3), strides=(2, 2), padding='same', name='conv_2_2')(relu_2)\nnorm_2_2 = BatchNormalization(name='norm_2_2')(conv_2_2)\nrelu_2_2 = Activation('relu', name='relu_2_2')(norm_2_2)\n\n# level three decomposition starts \nconv_b = Conv2D(filters=64, kernel_size=(3, 3), padding='same', name='conv_b')(input_l3)\nnorm_b = BatchNormalization(name='norm_b')(conv_b)\nrelu_b = Activation('relu', name='relu_b')(norm_b)\n\nconv_b_2 = Conv2D(128, kernel_size=(3, 3), padding='same', name='conv_b_2')(relu_b)\nnorm_b_2 = BatchNormalization(name='norm_b_2')(conv_b_2)\nrelu_b_2 = Activation('relu', name='relu_b_2')(norm_b_2)\n\n# concate level two and level three decomposition \nconcate_level_3 = concatenate([relu_2_2, relu_b_2])\nconv_3 = Conv2D(256, kernel_size=(3, 3), padding='same', name='conv_3')(concate_level_3)\nnorm_3 = BatchNormalization(name='nomr_3')(conv_3)\nrelu_3 = Activation('relu', name='relu_3')(norm_3)\n\nconv_3_2 = Conv2D(256, kernel_size=(3, 3), strides=(2, 2), padding='same', name='conv_3_2')(relu_3)\nnorm_3_2 = BatchNormalization(name='norm_3_2')(conv_3_2)\nrelu_3_2 = Activation('relu', name='relu_3_2')(norm_3_2)\n\n# level four decomposition start\nconv_c = Conv2D(64, kernel_size=(3, 3), padding='same', name='conv_c')(input_l4)\nnorm_c = BatchNormalization(name='norm_c')(conv_c)\nrelu_c = Activation('relu', name='relu_c')(norm_c)\n\nconv_c_2 = Conv2D(256, kernel_size=(3, 3), padding='same', name='conv_c_2')(relu_c)\nnorm_c_2 = BatchNormalization(name='norm_c_2')(conv_c_2)\nrelu_c_2 = Activation('relu', name='relu_c_2')(norm_c_2)\n\nconv_c_3 = Conv2D(256, kernel_size=(3, 3), padding='same', name='conv_c_3')(relu_c_2)\nnorm_c_3 = BatchNormalization(name='norm_c_3')(conv_c_3)\nrelu_c_3 = Activation('relu', name='relu_c_3')(norm_c_3)\n\n# concate level level three and level four decomposition\nconcate_level_4 = concatenate([relu_3_2, relu_c_3])\nconv_4 = Conv2D(256, kernel_size=(3, 3), padding='same', name='conv_4')(concate_level_4)\nnorm_4 = BatchNormalization(name='norm_4')(conv_4)\nrelu_4 = Activation('relu', name='relu_4')(norm_4)\n\nconv_4_2 = Conv2D(256, kernel_size=(3, 3), strides=(2, 2), padding='same', name='conv_4_2')(relu_4)\nnorm_4_2 = BatchNormalization(name='norm_4_2')(conv_4_2)\nrelu_4_2 = Activation('relu', name='relu_4_2')(norm_4_2)\n\nconv_5_1 = Conv2D(128, kernel_size=(3, 3), padding='same', name='conv_5_1')(relu_4_2)\nnorm_5_1 = BatchNormalization(name='norm_5_1')(conv_5_1)\nrelu_5_1 = Activation('relu', name='relu_5_1')(norm_5_1)\n\n\nx = Reshape((-1,128))(relu_5_1)\nx = Capsule(4,16, 3, True)(x)  \nx = Capsule(4, 16, 3, True)(x)\nx = Capsule(4, 16, 3, True)(x)\ncapsule=Capsule(2,16,3,True)(x)\n\n\noutputt = Lambda(lambda z: K.sqrt(K.sum(K.square(z), 2)))(capsule)\n\nResCapsnet_Binarymodel = Model(input_, outputs=outputt)\n\n\n","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:16.252007Z","iopub.execute_input":"2022-02-06T09:49:16.252434Z","iopub.status.idle":"2022-02-06T09:49:19.640433Z","shell.execute_reply.started":"2022-02-06T09:49:16.252399Z","shell.execute_reply":"2022-02-06T09:49:19.639691Z"},"trusted":true},"execution_count":17,"outputs":[{"name":"stderr","text":"2022-02-06 09:49:16.710518: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:16.806576: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:16.807346: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:16.808566: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX512F FMA\nTo enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n2022-02-06 09:49:16.809642: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:16.810345: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:16.810943: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:18.546031: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:18.546864: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:18.547548: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2022-02-06 09:49:18.548222: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 15403 MB memory:  -> device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:00:04.0, compute capability: 6.0\n","output_type":"stream"}]},{"cell_type":"code","source":"ResCapsnet_Binarymodel.summary()","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:19.642585Z","iopub.execute_input":"2022-02-06T09:49:19.643132Z","iopub.status.idle":"2022-02-06T09:49:19.672578Z","shell.execute_reply.started":"2022-02-06T09:49:19.643073Z","shell.execute_reply":"2022-02-06T09:49:19.671914Z"},"trusted":true},"execution_count":18,"outputs":[{"name":"stdout","text":"Model: \"model\"\n__________________________________________________________________________________________________\nLayer (type)                    Output Shape         Param #     Connected to                     \n==================================================================================================\nthe_input (InputLayer)          [(None, 224, 224, 3) 0                                            \n__________________________________________________________________________________________________\nwavelet (Lambda)                [(None, 112, 112, 12 0           the_input[0][0]                  \n__________________________________________________________________________________________________\nconv_1 (Conv2D)                 (None, 112, 112, 64) 6976        wavelet[0][0]                    \n__________________________________________________________________________________________________\nnorm_1 (BatchNormalization)     (None, 112, 112, 64) 256         conv_1[0][0]                     \n__________________________________________________________________________________________________\nrelu_1 (Activation)             (None, 112, 112, 64) 0           norm_1[0][0]                     \n__________________________________________________________________________________________________\nconv_1_2 (Conv2D)               (None, 56, 56, 64)   36928       relu_1[0][0]                     \n__________________________________________________________________________________________________\nconv_a (Conv2D)                 (None, 56, 56, 64)   6976        wavelet[0][1]                    \n__________________________________________________________________________________________________\nnorm_1_2 (BatchNormalization)   (None, 56, 56, 64)   256         conv_1_2[0][0]                   \n__________________________________________________________________________________________________\nnorm_a (BatchNormalization)     (None, 56, 56, 64)   256         conv_a[0][0]                     \n__________________________________________________________________________________________________\nrelu_1_2 (Activation)           (None, 56, 56, 64)   0           norm_1_2[0][0]                   \n__________________________________________________________________________________________________\nrelu_a (Activation)             (None, 56, 56, 64)   0           norm_a[0][0]                     \n__________________________________________________________________________________________________\nconcatenate (Concatenate)       (None, 56, 56, 128)  0           relu_1_2[0][0]                   \n                                                                 relu_a[0][0]                     \n__________________________________________________________________________________________________\nconv_2 (Conv2D)                 (None, 56, 56, 128)  147584      concatenate[0][0]                \n__________________________________________________________________________________________________\nconv_b (Conv2D)                 (None, 28, 28, 64)   6976        wavelet[0][2]                    \n__________________________________________________________________________________________________\nnorm_2 (BatchNormalization)     (None, 56, 56, 128)  512         conv_2[0][0]                     \n__________________________________________________________________________________________________\nnorm_b (BatchNormalization)     (None, 28, 28, 64)   256         conv_b[0][0]                     \n__________________________________________________________________________________________________\nrelu_2 (Activation)             (None, 56, 56, 128)  0           norm_2[0][0]                     \n__________________________________________________________________________________________________\nrelu_b (Activation)             (None, 28, 28, 64)   0           norm_b[0][0]                     \n__________________________________________________________________________________________________\nconv_2_2 (Conv2D)               (None, 28, 28, 128)  147584      relu_2[0][0]                     \n__________________________________________________________________________________________________\nconv_b_2 (Conv2D)               (None, 28, 28, 128)  73856       relu_b[0][0]                     \n__________________________________________________________________________________________________\nnorm_2_2 (BatchNormalization)   (None, 28, 28, 128)  512         conv_2_2[0][0]                   \n__________________________________________________________________________________________________\nnorm_b_2 (BatchNormalization)   (None, 28, 28, 128)  512         conv_b_2[0][0]                   \n__________________________________________________________________________________________________\nconv_c (Conv2D)                 (None, 14, 14, 64)   6976        wavelet[0][3]                    \n__________________________________________________________________________________________________\nrelu_2_2 (Activation)           (None, 28, 28, 128)  0           norm_2_2[0][0]                   \n__________________________________________________________________________________________________\nrelu_b_2 (Activation)           (None, 28, 28, 128)  0           norm_b_2[0][0]                   \n__________________________________________________________________________________________________\nnorm_c (BatchNormalization)     (None, 14, 14, 64)   256         conv_c[0][0]                     \n__________________________________________________________________________________________________\nconcatenate_1 (Concatenate)     (None, 28, 28, 256)  0           relu_2_2[0][0]                   \n                                                                 relu_b_2[0][0]                   \n__________________________________________________________________________________________________\nrelu_c (Activation)             (None, 14, 14, 64)   0           norm_c[0][0]                     \n__________________________________________________________________________________________________\nconv_3 (Conv2D)                 (None, 28, 28, 256)  590080      concatenate_1[0][0]              \n__________________________________________________________________________________________________\nconv_c_2 (Conv2D)               (None, 14, 14, 256)  147712      relu_c[0][0]                     \n__________________________________________________________________________________________________\nnomr_3 (BatchNormalization)     (None, 28, 28, 256)  1024        conv_3[0][0]                     \n__________________________________________________________________________________________________\nnorm_c_2 (BatchNormalization)   (None, 14, 14, 256)  1024        conv_c_2[0][0]                   \n__________________________________________________________________________________________________\nrelu_3 (Activation)             (None, 28, 28, 256)  0           nomr_3[0][0]                     \n__________________________________________________________________________________________________\nrelu_c_2 (Activation)           (None, 14, 14, 256)  0           norm_c_2[0][0]                   \n__________________________________________________________________________________________________\nconv_3_2 (Conv2D)               (None, 14, 14, 256)  590080      relu_3[0][0]                     \n__________________________________________________________________________________________________\nconv_c_3 (Conv2D)               (None, 14, 14, 256)  590080      relu_c_2[0][0]                   \n__________________________________________________________________________________________________\nnorm_3_2 (BatchNormalization)   (None, 14, 14, 256)  1024        conv_3_2[0][0]                   \n__________________________________________________________________________________________________\nnorm_c_3 (BatchNormalization)   (None, 14, 14, 256)  1024        conv_c_3[0][0]                   \n__________________________________________________________________________________________________\nrelu_3_2 (Activation)           (None, 14, 14, 256)  0           norm_3_2[0][0]                   \n__________________________________________________________________________________________________\nrelu_c_3 (Activation)           (None, 14, 14, 256)  0           norm_c_3[0][0]                   \n__________________________________________________________________________________________________\nconcatenate_2 (Concatenate)     (None, 14, 14, 512)  0           relu_3_2[0][0]                   \n                                                                 relu_c_3[0][0]                   \n__________________________________________________________________________________________________\nconv_4 (Conv2D)                 (None, 14, 14, 256)  1179904     concatenate_2[0][0]              \n__________________________________________________________________________________________________\nnorm_4 (BatchNormalization)     (None, 14, 14, 256)  1024        conv_4[0][0]                     \n__________________________________________________________________________________________________\nrelu_4 (Activation)             (None, 14, 14, 256)  0           norm_4[0][0]                     \n__________________________________________________________________________________________________\nconv_4_2 (Conv2D)               (None, 7, 7, 256)    590080      relu_4[0][0]                     \n__________________________________________________________________________________________________\nnorm_4_2 (BatchNormalization)   (None, 7, 7, 256)    1024        conv_4_2[0][0]                   \n__________________________________________________________________________________________________\nrelu_4_2 (Activation)           (None, 7, 7, 256)    0           norm_4_2[0][0]                   \n__________________________________________________________________________________________________\nconv_5_1 (Conv2D)               (None, 7, 7, 128)    295040      relu_4_2[0][0]                   \n__________________________________________________________________________________________________\nnorm_5_1 (BatchNormalization)   (None, 7, 7, 128)    512         conv_5_1[0][0]                   \n__________________________________________________________________________________________________\nrelu_5_1 (Activation)           (None, 7, 7, 128)    0           norm_5_1[0][0]                   \n__________________________________________________________________________________________________\nreshape (Reshape)               (None, 49, 128)      0           relu_5_1[0][0]                   \n__________________________________________________________________________________________________\ncapsule (Capsule)               (None, 4, 16)        8192        reshape[0][0]                    \n__________________________________________________________________________________________________\ncapsule_1 (Capsule)             (None, 4, 16)        1024        capsule[0][0]                    \n__________________________________________________________________________________________________\ncapsule_2 (Capsule)             (None, 4, 16)        1024        capsule_1[0][0]                  \n__________________________________________________________________________________________________\ncapsule_3 (Capsule)             (None, 2, 16)        512         capsule_2[0][0]                  \n__________________________________________________________________________________________________\nlambda (Lambda)                 (None, 2)            0           capsule_3[0][0]                  \n==================================================================================================\nTotal params: 4,437,056\nTrainable params: 4,432,320\nNon-trainable params: 4,736\n__________________________________________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"code","source":"X_val.shape","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:19.674947Z","iopub.execute_input":"2022-02-06T09:49:19.675164Z","iopub.status.idle":"2022-02-06T09:49:19.682643Z","shell.execute_reply.started":"2022-02-06T09:49:19.675140Z","shell.execute_reply":"2022-02-06T09:49:19.681934Z"},"trusted":true},"execution_count":19,"outputs":[{"execution_count":19,"output_type":"execute_result","data":{"text/plain":"(300, 224, 224, 3)"},"metadata":{}}]},{"cell_type":"code","source":"X_train.shape","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:19.684035Z","iopub.execute_input":"2022-02-06T09:49:19.684298Z","iopub.status.idle":"2022-02-06T09:49:19.691584Z","shell.execute_reply.started":"2022-02-06T09:49:19.684265Z","shell.execute_reply":"2022-02-06T09:49:19.690847Z"},"trusted":true},"execution_count":20,"outputs":[{"execution_count":20,"output_type":"execute_result","data":{"text/plain":"(2700, 224, 224, 3)"},"metadata":{}}]},{"cell_type":"code","source":"!free -m","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:19.692643Z","iopub.execute_input":"2022-02-06T09:49:19.693426Z","iopub.status.idle":"2022-02-06T09:49:20.430838Z","shell.execute_reply.started":"2022-02-06T09:49:19.693307Z","shell.execute_reply":"2022-02-06T09:49:20.430043Z"},"trusted":true},"execution_count":21,"outputs":[{"name":"stdout","text":"              total        used        free      shared  buff/cache   available\nMem:          16009        8577        3594           8        3837        7148\nSwap:             0           0           0\n","output_type":"stream"}]},{"cell_type":"code","source":"del X, Y","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:20.433430Z","iopub.execute_input":"2022-02-06T09:49:20.433736Z","iopub.status.idle":"2022-02-06T09:49:20.493740Z","shell.execute_reply.started":"2022-02-06T09:49:20.433690Z","shell.execute_reply":"2022-02-06T09:49:20.491844Z"},"trusted":true},"execution_count":22,"outputs":[]},{"cell_type":"code","source":"import gc\ngc.collect()","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:20.496413Z","iopub.execute_input":"2022-02-06T09:49:20.496604Z","iopub.status.idle":"2022-02-06T09:49:20.720629Z","shell.execute_reply.started":"2022-02-06T09:49:20.496581Z","shell.execute_reply":"2022-02-06T09:49:20.719907Z"},"trusted":true},"execution_count":23,"outputs":[{"execution_count":23,"output_type":"execute_result","data":{"text/plain":"11222"},"metadata":{}}]},{"cell_type":"code","source":"opt= Adam()\nResCapsnet_Binarymodel.compile(\n    optimizer=opt,\n    loss=margin_loss,\n    metrics=['accuracy'],\n)\nfrom tensorflow.keras.callbacks import EarlyStopping\nfrom tensorflow.keras.callbacks import ModelCheckpoint\n\n\n\nes =  ModelCheckpoint(monitor='val_accuracy', mode='max', filepath='./model.h5',save_weights_only=True,save_best_only=True  )\nes1 =  ModelCheckpoint(monitor='val_loss', mode='min', filepath='./model_loss.h5',save_weights_only=True,save_best_only=True  )\n\nhistory=ResCapsnet_Binarymodel.fit(X_train,Y_train ,epochs=20,validation_split=0.10, batch_size=32, callbacks=[es,es1])","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:49:20.721939Z","iopub.execute_input":"2022-02-06T09:49:20.722202Z","iopub.status.idle":"2022-02-06T09:51:51.094757Z","shell.execute_reply.started":"2022-02-06T09:49:20.722167Z","shell.execute_reply":"2022-02-06T09:51:51.093936Z"},"trusted":true},"execution_count":24,"outputs":[{"name":"stderr","text":"2022-02-06 09:49:21.297331: W tensorflow/core/framework/cpu_allocator_impl.cc:80] Allocation of 1463132160 exceeds 10% of free system memory.\n2022-02-06 09:49:22.873932: W tensorflow/core/framework/cpu_allocator_impl.cc:80] Allocation of 1463132160 exceeds 10% of free system memory.\n2022-02-06 09:49:23.943381: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n","output_type":"stream"},{"name":"stdout","text":"Epoch 1/20\n","output_type":"stream"},{"name":"stderr","text":"2022-02-06 09:49:29.979146: I tensorflow/stream_executor/cuda/cuda_dnn.cc:369] Loaded cuDNN version 8005\n","output_type":"stream"},{"name":"stdout","text":"76/76 [==============================] - 20s 104ms/step - loss: 0.1627 - accuracy: 0.6066 - val_loss: 0.1238 - val_accuracy: 0.4963\nEpoch 2/20\n76/76 [==============================] - 6s 74ms/step - loss: 0.0683 - accuracy: 0.8000 - val_loss: 0.3244 - val_accuracy: 0.4963\nEpoch 3/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0556 - accuracy: 0.8235 - val_loss: 0.0842 - val_accuracy: 0.7852\nEpoch 4/20\n76/76 [==============================] - 6s 74ms/step - loss: 0.0335 - accuracy: 0.9008 - val_loss: 0.1460 - val_accuracy: 0.6148\nEpoch 5/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0345 - accuracy: 0.8984 - val_loss: 0.1000 - val_accuracy: 0.6889\nEpoch 6/20\n76/76 [==============================] - 6s 75ms/step - loss: 0.0280 - accuracy: 0.9226 - val_loss: 0.0547 - val_accuracy: 0.8222\nEpoch 7/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0200 - accuracy: 0.9436 - val_loss: 0.0284 - val_accuracy: 0.9185\nEpoch 8/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0157 - accuracy: 0.9547 - val_loss: 0.1168 - val_accuracy: 0.6185\nEpoch 9/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0131 - accuracy: 0.9642 - val_loss: 0.0765 - val_accuracy: 0.7852\nEpoch 10/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0079 - accuracy: 0.9778 - val_loss: 0.0064 - val_accuracy: 0.9889\nEpoch 11/20\n76/76 [==============================] - 6s 75ms/step - loss: 0.0085 - accuracy: 0.9798 - val_loss: 0.0242 - val_accuracy: 0.9259\nEpoch 12/20\n76/76 [==============================] - 6s 74ms/step - loss: 0.0061 - accuracy: 0.9848 - val_loss: 0.0076 - val_accuracy: 0.9815\nEpoch 13/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0060 - accuracy: 0.9840 - val_loss: 0.0542 - val_accuracy: 0.8630\nEpoch 14/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0047 - accuracy: 0.9889 - val_loss: 0.0080 - val_accuracy: 0.9815\nEpoch 15/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0068 - accuracy: 0.9815 - val_loss: 0.0113 - val_accuracy: 0.9704\nEpoch 16/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0046 - accuracy: 0.9864 - val_loss: 0.0354 - val_accuracy: 0.9074\nEpoch 17/20\n76/76 [==============================] - 6s 75ms/step - loss: 0.0034 - accuracy: 0.9918 - val_loss: 0.0634 - val_accuracy: 0.8407\nEpoch 18/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0030 - accuracy: 0.9934 - val_loss: 0.0057 - val_accuracy: 0.9852\nEpoch 19/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0110 - accuracy: 0.9687 - val_loss: 0.0206 - val_accuracy: 0.9333\nEpoch 20/20\n76/76 [==============================] - 6s 73ms/step - loss: 0.0040 - accuracy: 0.9885 - val_loss: 0.0216 - val_accuracy: 0.9407\n","output_type":"stream"}]},{"cell_type":"code","source":"\n\nplt.plot(history.history['loss'])\nplt.plot(history.history['val_loss'])\nplt.title('Loss plot')\nplt.ylabel('loss')\nplt.xlabel('epoch')\nplt.legend(['train', 'val'], loc='upper right')\nplt.show()\n\n","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:51:51.099663Z","iopub.execute_input":"2022-02-06T09:51:51.099881Z","iopub.status.idle":"2022-02-06T09:51:51.300377Z","shell.execute_reply.started":"2022-02-06T09:51:51.099855Z","shell.execute_reply":"2022-02-06T09:51:51.299713Z"},"trusted":true},"execution_count":25,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"plt.plot(history.history['accuracy'])\nplt.plot(history.history['val_accuracy'])\nplt.title('Model Accuracy')\nplt.ylabel('accuracy')\nplt.xlabel('epoch')\nplt.legend(['train', 'val'], loc='lower right')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:51:51.301491Z","iopub.execute_input":"2022-02-06T09:51:51.301751Z","iopub.status.idle":"2022-02-06T09:51:51.495944Z","shell.execute_reply.started":"2022-02-06T09:51:51.301717Z","shell.execute_reply":"2022-02-06T09:51:51.495318Z"},"trusted":true},"execution_count":26,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"ResCapsnet_Binarymodel.load_weights('./model_loss.h5')\npred = ResCapsnet_Binarymodel.predict(X_val)\nResCapsnet_Binarymodel.evaluate(X_val,Y_val)","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:58:30.230115Z","iopub.execute_input":"2022-02-06T09:58:30.230683Z","iopub.status.idle":"2022-02-06T09:58:31.717678Z","shell.execute_reply.started":"2022-02-06T09:58:30.230644Z","shell.execute_reply":"2022-02-06T09:58:31.716857Z"},"trusted":true},"execution_count":34,"outputs":[{"name":"stdout","text":"10/10 [==============================] - 0s 27ms/step - loss: 0.0074 - accuracy: 0.9867\n","output_type":"stream"},{"execution_count":34,"output_type":"execute_result","data":{"text/plain":"[0.00736577995121479, 0.9866666793823242]"},"metadata":{}}]},{"cell_type":"code","source":"confusion_matrix=sklearn.metrics.confusion_matrix(Y_val.argmax(axis=1), ResCapsnet_Binarymodel.predict(X_val).argmax(axis=1))\nax= plt.subplot()\nsns.set(font_scale=2.0)\nsns.heatmap(confusion_matrix, annot=True, fmt='g', ax=ax);  #annot=True to annotate cells, ftm='g' to disable scientific notation\n\n# labels, title and ticks\nax.set_xlabel('Predicted labels',fontdict={'size':'18'});ax.set_ylabel('True labels',fontdict={'size':'18'}); \nax.set_title('Confusion Matrix Capsule',fontdict={'size':'24'}); \nax.tick_params(axis='both', which='major', labelsize=20)\nax.xaxis.set_ticklabels(['Non-Tumor', 'Tumor']); ax.yaxis.set_ticklabels(['Non-Tumor', 'Tumor']);","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:58:34.508874Z","iopub.execute_input":"2022-02-06T09:58:34.509725Z","iopub.status.idle":"2022-02-06T09:58:35.419490Z","shell.execute_reply.started":"2022-02-06T09:58:34.509687Z","shell.execute_reply":"2022-02-06T09:58:35.418828Z"},"trusted":true},"execution_count":35,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"from sklearn.metrics import classification_report\nprint(classification_report(Y_val.argmax(axis=1), pred.argmax(axis=1), target_names=['Non-Tumor', 'Tumor']))","metadata":{"execution":{"iopub.status.busy":"2022-02-06T09:58:40.252584Z","iopub.execute_input":"2022-02-06T09:58:40.253250Z","iopub.status.idle":"2022-02-06T09:58:40.265471Z","shell.execute_reply.started":"2022-02-06T09:58:40.253121Z","shell.execute_reply":"2022-02-06T09:58:40.264606Z"},"trusted":true},"execution_count":36,"outputs":[{"name":"stdout","text":"              precision    recall  f1-score   support\n\n   Non-Tumor       0.97      1.00      0.99       150\n       Tumor       1.00      0.97      0.99       150\n\n    accuracy                           0.99       300\n   macro avg       0.99      0.99      0.99       300\nweighted avg       0.99      0.99      0.99       300\n\n","output_type":"stream"}]}]}