Ask HN: Is it enough to use deep learning?
Looks like deep learning models are good enough to solve almost all types of ML problems. So isn't learning that sufficient instead of doing every other ML algorithms? Or others are still needed in specific situations?