Hacking in Bloomington!
I’ve been helping Eric H a bit on a paper about training classifiers for webspam detection (sort of a “how to do machine learning for your security problem” paper), and I was supposed to write about Naive Bayes classifiers. It occurred to me that I’d never implemented one.
So I did! It’s straightforward, because naive classifier is naive.
Also, I did a bit of hacking on Snippets, which is like weekly atlhack notes for people’s non-side projects. It’s up and running, and some people at Indiana are even using it! As of last night, it sends weekly reminders to people, so they’ll remember to update (these are opt-in).
*hugs from the frozen north*