Select all true statements about Convolutional Neural Network $($CNN$)?0$ point if not a perfect

match.

There are a POOL layers after several convolutional layers

The deeper the CNN$,$ the lower training error is

The first several layers and the middle layers can also be the fully connected layers

There are only two types of layers $($convolutional and fully connected layers$)$ in CNN

A convolutional layer is placed after several POOL layers

The number of layers in CNN must be three, which includes convolutional layer, pooling

layer, and fully connected layer

The last few layers are the fully connected layers

The primary goal of using Convolutional layer is to reduce the number of parameters in

CNN

Applying $22$ pooling with stride $2$ on the $1010$ layer will result in $44$ output

Convolving an input of size $777716$ with $32$ filters of size $77$ using stride of $2$ and no

padding will result in output of size $363632$

Pooling layers are parameterless, so they have no impact on the backpropagation

CNN can be used to find the pattern in image and is useful for object recognition

The relationship between Convolutional Neural Network $($CNN$)$ and Cable News

Network $($CNN$)$ is the later has builded the former

All sentences $1-13$ are true

All sentences $1-13$ are false