TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

A Theory of Universal Learning [pdf]

93 pointsby emreover 4 years ago

1 comment

bra-ketover 4 years ago
Calling this &#x27;universal learning&#x27; is a stretch (and in HN speak &#x27;click-bait&#x27;). The paper only talks about a particular subclass of learning algorithms in supervised learning domain within PAC framework.<p>Specially they talk about learning algorithms that minimize some error function from training examples. That&#x27;s not how learning happens in living organisms. Therefore calling it &#x27;universal&#x27; is an unwarranted generalization, kind of like &#x27;theory of everything&#x27;.<p>A more appropriate name would be &#x27;Distribution-dependent Supervised PAC Learning&#x27; or something along those lines. It&#x27;s a solid work which addresses a particular niche of a particular theory.
评论 #25041163 未加载
评论 #25042144 未加载