Question $10$

$6pts$

The following code was used to design a NN model to classify the CIFAR$10$ dataset. We used a $50\%$ dropout in the fully connected layers. The summary of the model is shown in the image below. Fill in the blanks.

#defining the model

model$\_$a$=$Sequential$()$

model$\_$a$.$add$($Conv$2$D$(,(3,3),$ padding$=$'same',

input$\_$shape$=$x$\_$train.shape$[1$:$]))$

model$\_$a$.$add$($Activation$(\text{'}$relu$\text{'}))$

model$\_$a$.$add$($Conv$2$D$($ model$\_$a$.$add$($Activation$(\text{'}$relu$\text{'}))$

model$\_$a$.$add$($MaxPooling$2$D$($pool$\_$size$=($

model$\_$a$.$add$($Conv$2$Di $(3,3),$ padding$=$'same'$))$ model$\_$a$.$add$($Activation$(\text{'}$relu$\text{'}))$

model$\_$a$.$add$($Conv$2$D $(3,3),$ padding$=$'same'$))$

model$\_$a$.$add$($Activation$(\text{'}$relu$\text{'}))$

model$\_$a$.$add$($MaxPooling$2$D$($pool$\_$size$=1$

$))$

model$\_$a$.$add$($Flatten$())$

model$\_$a$.$add$($Dense$($units$=$ activation$=$'relu'$))$

model$\_$a$.$add$($Dropout$($ model$\_$a$.$add$($Dense$($units$=,$ activation $=\text{'}))$ model$\_$a$.$summary$()$