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.

Bayesian Logic Programming

125 pointsby jlturnerover 9 years ago

9 comments

cedover 9 years ago
It&#x27;s by Stuart Russell&#x27;s group at Berkeley (he coauthored the AI bible with Peter Norvig). Their tutorial: <a href="http:&#x2F;&#x2F;bayesianlogic.github.io&#x2F;download&#x2F;BLOG-tutorial-2014.pdf" rel="nofollow">http:&#x2F;&#x2F;bayesianlogic.github.io&#x2F;download&#x2F;BLOG-tutorial-2014.p...</a> Page 58 has some good sample code. Semantics are on page 70.<p><i>Every well-formed BLOG model specifies a unique proper probability distribution over all possible worlds definable given its vocabulary •No infinite receding ancestor chains; •no conditioned cycles; •all expressions finitely evaluable; •Functions of countable sets</i><p>They instantiate some parts of the network and do inference with MCMC. I wonder how it compares to the Markov Logic approach from the University of Washington.
chmulligover 9 years ago
Interesting. They talk about plain old Metropolis Hastings, which is pretty questionable.<p>Anyone excited about this, I highly recommend checking out Stan; it&#x27;s under active development, actually works with real problems, and is used in the real world. With NUTS and HMC they&#x27;ve really made good on their promises, and quite soon they&#x27;ll have meaningful ADVI support. See this former discussion: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=10244771" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=10244771</a>
评论 #10758550 未加载
评论 #10757518 未加载
nlover 9 years ago
They really called their language &quot;blog&quot;? I have to think that wasn&#x27;t the best name ever..
评论 #10757312 未加载
nickpsecurityover 9 years ago
&quot;We&#x27;ve coded up the application that will run your business. It has a 80% chance of working correctly roughly 20% of the time with a 95% confidence interval.&quot;
评论 #10757242 未加载
评论 #10756522 未加载
emmanueloga_over 9 years ago
More resources related to this subject:<p><a href="http:&#x2F;&#x2F;probabilistic-programming.org" rel="nofollow">http:&#x2F;&#x2F;probabilistic-programming.org</a><p>There&#x27;s <i>probably</i> some alternative, actively developed projects that have the same objective as BLOG listed on that page.
argonautover 9 years ago
Doesn&#x27;t seem to be under active development. <a href="https:&#x2F;&#x2F;github.com&#x2F;BayesianLogic&#x2F;blog" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;BayesianLogic&#x2F;blog</a>
slyrusover 9 years ago
Nice ungooglable project name
probinsoover 9 years ago
also check out FIGARO and Hakaru.
florinutzover 9 years ago
why would darpa fund this?
评论 #10758162 未加载