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

And here I was expecting things like 'overfitting' and 'not having a holdout set or at least crossvalidating'.


You mean there are people whose machine learning homework in college/grad-school didn't threaten a zero for failing to cross-validate?

Well damn. What did I spend all those late nights in front of a bleak MATLAB environment for, some kind of best practices?! </tongue in cheek>


Yes, I didn't find this helpful at all. Also, it needs editing.


Hackernews is moving towards Yahoo news.




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

Search: