Why not also predict the error of the error prediction, and correct for that too... | Hacker News

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

		ttul on Oct 6, 2020 \| parent \| context \| favorite \| on: Gradient Boosted Decision Trees Why not also predict the error of the error prediction, and correct for that too? /s

DSingularity on Oct 7, 2020 [–]

You mock, but overfitting is is how most results are realized.

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