*>but to characterize our understanding of machine learning as not going beyond ... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

gambler on Feb 14, 2019 | parent | context | favorite | on: Neural Networks Are Essentially Polynomial Regress...

>but to characterize our understanding of machine learning as not going beyond 3 or 4 neurons is just plain crazy.

There is a difference between understanding how something works and understanding why it works. Completely different levels of predictive power.

dnautics on Feb 15, 2019 [–]

I'm sorry, lstms and convnets would seem to represent a rather high level of predictive power, given that someone postulated that they would work, and it turned out they were correct.

Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4
Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact