Describe the backpropagation algorithm and gradient descent method. $2)2-$D convolution. Given the input and two kernels as shown in the following figure:

compute the feature maps by applying kernel K$1$ and k$2($stride is defined as $1$ and zeros are padded around I$),$ respectively; b$.$ Re$-$compute the feature map if the stride is $2$; c$.$ Compute the new feature maps after we apply a downsampling process, max polling, and average polling $($filter size is $2$ x $2),$ respectively, to the obtained feature maps in $($a$).$