Because when we sit down and think about a problem, when we take the time to not only understand what our feature space “is” and what it “implies” in the real-world — then we are acting like machine learning scientists. Otherwise, we [are] just a bunch of machine learning engineers, blindly performing black box learning and operating a set of R, MATLAB, and Python libraries.
The takeaway is this: machine learning isn’t a tool. It’s a methodology with a rational thought process that is entirely dependent on the problem we are trying to solve.
Get off the deep learning bandwagon and get some perspective – PyImageSearch
Kenkyuu 