Convolution Neural Networks (CNNs) in essence are neural networks that employ the convolution operation (instead of a fully connected layer) as one of its layers. CNNs are an incredibly successful technology that has been applied to problems wherein the input data on which predictions are to be made has a known grid like topology like a time series (which is a 1-D grid) or an image (which is a 2-D grid).
Convolution Operation
Let us start developing intuition for the convolution operation in one dimension. Given an input I(t) and a kernel K(a) the convolution operation ...