[Question] How do you do hyperparameter searches in ML?

I know how to do hy­per­pa­ram­e­ter searches. ☺

This is a sur­vey. I want to know is how you do hy­per­pa­ram­e­ter searches. It doesn’t mat­ter whether your sys­tem is good or bad. I won’t judge you. I just want to know what sys­tems other peo­ple are us­ing in the real world right now.

Any in­for­ma­tion you’re will­ing to share would help me out here, but there are two ques­tions I’m es­pe­cially in­ter­ested in.

  1. What al­gorithm do you use? (Do you use ran­dom search, grid search or Bayes search? Do you do some iter­a­tive pro­cess? Do you do some­thing else en­tirely?)

  2. Do you cache any­thing? If so, what’s your pro­cess?

I’m also cu­ri­ous what in­dus­try you’re in, but if you’re not com­fortable shar­ing that some in­for­ma­tion is bet­ter than none.

No comments.