Benefits of depth in neural networks — mentioned by j2kun

Thu, 07 Apr 2016 17:39:26 GMT

I spent a few weeks closely reading this book and I have to disagree with the majority here. I didn't like the book at all. And I am an advanced math geek. My main issue is that the book tells you all about the different parameter tweaks, but passes little concrete wisdom to the reader. It doesn't distinguish between modeling assumptions, and it replaces very simple explanations of concepts with complicated paragraphs that I can't make sense of. I think it boils down to something that I have been feeling and hearing a lot in the past few years: the statistical jargon is so overwhelming that…

Paper and discussion history. A mention is not an endorsement.

Benefits of depth in neural networks · HackerPapers discussions

Benefits of depth in neural networks — mentioned by j2kun