The main lecture is in day2 (sorry I can't count). day2.pdf shows the full tutorials with annotations, but is a pain to read except in presentation mode. The day2_prn versions don't show the annotations, but are easier to read and to print. The _prn_4 and prn_8 version show 4 and 8 pages per day

I've added some of my lecture notes on SVMs. This takes me three lectures and covers deriving the formula for the maximum margin hyperplane (supportVec), an explanation for constrained optimisation and the dual problem (dualAlgorithm) and some properties of kernels (kernels)

A good introduction to graphical models is A Tutorial on Learning With Bayesian Networks by David Heckerman