Had lots of fun on AI class today. Learned a lot of neat stuff about
neural networks. I concluded that a non-supervised neural network
could probably help me sort my emails in a nice way. Have to write an
article on "Gene searching using neural networks". In the meantime,
some Bayesian madness to melt your brain. I actually understood
98.723% of this one.

And now I think I understood why Bayesian Learning has so much hype
around it.

I was reading a paper on gene recognition which suggestes that
combining different methods may yield to significantly better
results. I feel that this is a very interesting field. Maybe I should
choose this as the theme for my graduation thesis.

Henry Stern has done some intersting work on SpamAssassin. And
this guy knows who he’s fighting.

From the little I know about neural networks, and after reading some
papers on gene searching, it seems that using some ideas from genetics
could improve even more spam recognition. I could be very wrong, as
usual :)