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.

Model-Based Machine Learning

221 pointsby seycombiover 8 years ago

4 comments

oergiRover 8 years ago
The &quot;model&quot; in the title is the model of the world, as a probabilistic model. The good thing about such a model is that it explicitly states your beliefs about the world. Once you&#x27;ve defined it, in theory reasoning about it is straightforward. (In practice a lot of papers get written about how to do approximate inference.) It&#x27;s also straightforward to do unsupervised learning.<p>This is a different perspective from (most uses of) neural networks, which do not have this clear separation between the model and how to reason about it. It&#x27;s funny that Chris Bishop in 1995 wrote the textbook &quot;Neural Networks for Pattern Recognition&quot; and now is effectively arguing against using neural networks.<p>You can use both by using neural networks as &quot;factors&quot; (the black squares) in probabilistic models.
评论 #13102471 未加载
评论 #13102814 未加载
ThePhysicistover 8 years ago
I have to say the layout of this website looks great! Very accessible and clean. Was it made with a specific framework?
评论 #13100481 未加载
nkozyraover 8 years ago
I&#x27;ve never heard supervised learning referred to as model-based learning.
评论 #13100312 未加载
评论 #13100297 未加载
评论 #13100149 未加载
评论 #13100132 未加载
geoooooooooboxover 8 years ago
Anybody know if Scala&#x27;s Figaro software is in the same category as Church?
评论 #13101798 未加载