What are two reasons to use a pooling layer in a Convolution Neural Network (CNN)? (Choose two)
to increase granularity of information
to make the feature map invariant to transformation
to extract features of different spatial scale
to extract a summarized value from an area of the image
Submit