As Brett mentioned, there's multiple aspects to machine learning. I really don't have much experience, but the one book I've seen used in nearly every grad level machine learning course is: http://www.amazon.com/Machine-Learning-Tom-M-Mitchell/dp/0070428077/ref=sr_1_1?ie=UTF8&s=books&qid=1207269886&sr=1-1 ...it doesn't have all too much depth in any area, and covered only a little of SVMs (which are kinda the current 'state of the art' model), but it provides you a good base for where the field began, the core things that have stuck with the field, and why those things work.
recommendation on book for learning about exterior calculus and DEC
2008-03-31: dyn systems professor ext calc references
Arnold: Mathematicla methods of classical mechanics
email from professor
Learning CUDA
2008-04-12: copy lecture notes to my ipod
see if the library hold training classes on how to do literature searches
ask Torrellas about this: what if the hardware could check for cancellation errors? it could squash and rollback.
--: Talk to Anil about cardiac stuff
--: order the Atkinson book on Introduction to Numerical Analysis.
--: plan out next actions on all these project ideas