Max Pooling Forward

Problem

Forward pass of 2D max pooling over a square matrix using a k x k window and the given stride, without padding.

The signature is max_pool(x, k, stride). The output size per dimension is (n - k) // stride + 1.

Hint

Initialize the running maximum from the window's top-left element, not from zero.

case 1

in: [[[1,2,3,4],[5,6,7,8],[9,10,11,12],[13,14,15,16]],2,2]

expected: [[6,8],[14,16]]

case 2

in: [[[1,2,3],[4,5,6],[7,8,9]],2,1]

expected: [[5,6],[8,9]]

case 3

in: [[[1,2,3],[4,5,6],[7,8,9]],3,3]

expected: [[9]]

case 4

in: [[[-1,-2],[-3,-4]],2,2]

expected: [[-1]]

Starter codePython

def max_pool(x, k, stride):
    # Your code here
    pass

4 test cases · run your code against them in the editor