Code Assignment

?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

def HappyModel(input_shape):     X_input = Input(input_shape)           X = ZeroPadding2D((3,3))(X_input)     X = Conv2D(18,(7,7),strides=(1,1),name="conv0")(X)     X = BatchNormalization(axis=3, name="bn0")(X)     X = Activation("relu")(X)           X = MaxPooling2D((2,2), name="max_pool")(X)           X = Flatten()(X)     X = Dense(1,activation="sigmoid", name="fC")(X)           model = Model(input = X_input, outputs = X, name="happy model")      return model   happyModel = HappyModel(X_train.shape[1:])        happyModel.compile(optimizer="adam", loss="binary_crossentropy", metrics=["accuracy"])       happyModel.fit(x=X_train, y=Y_train,epochs=10, batch_size=20)       preds = happyModel.evaluate(x=X_test,y=Y_test)   img_path = 'images/smile.jpg'     img = image.load_img(img_path, target_size=(64, 64)) imshow(img)   x = image.img_to_array(img) x = np.expand_dims(x, axis=0) x = preprocess_input(x) print(happyModel.predict(x))