[FoRK] Bayesian networks without tears

Jeff Bone jbone at place.org
Thu Aug 28 20:20:48 PDT 2008


For folks whose eyes glaze over when people start mentioning things  
like "Bayesian networks" here's an excellent introduction that isn't  
terribly math-heavy and doesn't require much of a grounding in  
probability or stats:

   http://www.cs.ubc.ca/~murphyk/Bayes/Charniak_91.pdf

I'd read this some time ago but it serendipitously popped up on reddit  
(IIRC) a couple of days ago...  I was trying to explain (as it seems  
I'm constantly trying to explain) the difference between e.g. Bayesian  
nets and neural nets to somebody else (offlist) the other day, and why  
one is entirely "transparent" and the other not necessarily so, and  
when I saw the post it sparked my recollection of reading this back  
when it was published.

At any rate, recommended if you're interested in such things.

jb



More information about the FoRK mailing list