Hacker News new | past | comments | ask | show | jobs | submit login

Yeah, backprop gives efficient gradient calcs to train at a given net architecture. To find the best architecture though people try various other methods, eg random search or Gaussian processes, which don’t evaluate derivatives wrt architecture Params.



Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: