Question

Why we have to do training of neural network after feature extraction?

Answer 1

Answer:

Convolution neural network you can think in different angle, where architecture of the network provides a technique for reduction of input space dimension, retaining important features to be learned easily.