So I want to classify the images. I am confused on how to seperate the 3 RGB channels and input into a logistic regression model. I only know of 2D linear data which is fairly easy.