Yioop - PHP Search Engine

-- Nov 2 In-Class Exercise Thread

Resource Description for Screen Shot 2021-11-07 at 1.06.41 AM.png

Input m: 64 * 64 = 4096 Output n: No. of filters * Output shape = 32 * (60 * 60) = 115200 Therefore, using Glorot and Bengio initialization Wi,j~U(-(6/(4096+115200))^0.5,(6(/4096+115200))^0.5) Wi,j~U(−0.0071,0.0071)

(Edited: 2021-11-07)

((resource:Screen Shot 2021-11-07 at 1.06.41 AM.png|Resource Description for Screen Shot 2021-11-07 at 1.06.41 AM.png)) Input m: 64 * 64 = 4096 Output n: No. of filters * Output shape = 32 * (60 * 60) = 115200 Therefore, using Glorot and Bengio initialization Wi,j~U(-(6/(4096+115200))^0.5,(6(/4096+115200))^0.5) Wi,j~U(−0.0071,0.0071)

-- Nov 2 In-Class Exercise Thread

model.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', input_shape=(64, 64, 1)))

m = 64*64 = 4096 n = 64*64*32 = 131072

Resource Description for WhatsApp Image 2021-11-08 at 4.38.52 PM (1).jpeg

(Edited: 2021-11-08)

model.add(Conv2D(32, (3, 3), activation='relu', kernel_initializer='he_uniform', input_shape=(64, 64, 1))) m = 64*64 = 4096 n = 64*64*32 = 131072 ((resource:WhatsApp Image 2021-11-08 at 4.38.52 PM (1).jpeg|Resource Description for WhatsApp Image 2021-11-08 at 4.38.52 PM (1).jpeg))

-- Nov 3 In-Class Exercise Thread

((resource:nov3.jpeg|Resource Description for nov3.jpeg))

-- Nov 3 In-Class Exercise Thread

Inputs to first layer, m: 64 * 64 = 4096

Outputs from first layer, n: No. of filters * Output shape = 32 * (64 * 64)

Using Glorot and Bengio initialization

(Edited: 2021-11-09)

Inputs to first layer, m: 64 * 64 = 4096 Outputs from first layer, n: No. of filters * Output shape = 32 * (64 * 64) Using Glorot and Bengio initialization ((resource:unnamed.jpg|Resource Description for unnamed.jpg))

-- Nov 3 In-Class Exercise Thread

Resource Description for AI.jpeg

((resource:AI.jpeg|Resource Description for AI.jpeg))