Can someone help me understand when you'd want to use neural networks vs multivariate regressions? I know there is some obvious reasoning: multivariate regression will give you a formula where NNs are entirely "blackbox". But can NNs uncover patterns that mutivariate cannot? My hunch is yes, but I can't conceptualize why.
I'm a novice data scientist and have Googled quite a bit on this question, but can never wrap my head around the answer. I'm hoping some of you can point me the right direction.
[link][2 comments]