MetaChat REGISTER   ||   LOGIN   ||   IMAGES ARE OFF   ||   RECENT COMMENTS




artphoto by splunge
artphoto by TheophileEscargot
artphoto by Kronos_to_Earth
artphoto by ethylene

Home

About

Search

Archives

Mecha Wiki

Metachat Eye

Emcee

IRC Channels

IRC FAQ


 RSS


Comment Feed:

RSS

07 December 2010

An intuitive introduction to Bayesian reasoning. So I've been working through this -- slowly -- for the past couple of days, and I have to say I've found unexpectedly illuminating and rewarding.[More:]I know it's not exactly fun and this author is rather dry, but it seems he'd much rather you truly understand than just memorize a set of rules, and so he takes a lot of time to both step you through it and challenge you to figure out the next step on your own. In the end, he ties it up in a philosophical bow that seems to have significant implications for both the scientific method and artificial intelligence. Much more fascinating than I thought it would be.
Having just taught a statistics course, I got it right with 15% of the doctors.
posted by Obscure Reference 07 December | 14:07
Oh, that Yudkowsky guy scares me. And not in any way because of Bayesian reasoning.
posted by Madamina 07 December | 18:43
Madamina, I kind of agree, though I'm have a hard time putting my finger on it. Why does he scare you?

Here's my take:

My first impression of reading through the Less Wrong site is that it felt kind of cultish -- that he'd written all the core material didn't help. And there's something about his writing style that comes across (at least at first) as quite coldly arrogant, but also uncannily intimate.

But then I learned the core material consists merely of interrelated blog posts, not of some sort of bizarre systematic theological exegesis, and that he's very consciously trying to walk the walk he talks (e.g., learning to cultivate rationality or clarity) and incorporate others into the conversation about what that means -- and that made me feel a little less wary.

In the end though, reading this article really taught me a lot -- it did what it said on the package, and gave me a genuinely intuitive grasp of Bayes's Theorem. And that feels really empowering and unexpectedly exciting.
posted by treepour 07 December | 21:51
New Year's Resolutions || Nearly a decade ago,

HOME  ||   REGISTER  ||   LOGIN