Links to Stanford Machine Learning lectures, login not required

I'm sure you all know about www.ml-class.org - if I were to distillate all lectures to bare minimum it would be to learn about support vector machines (SVM).

Contrary to common believes SVMs are a VERY large class of learners. Linear SVM would be just 1 of 10^10 ways of using SVM. It can be used to do regression, classification, clustering, feature selection, feature weighting, feature engineering, projections, and visualizations. That's why I decided to post few links to the prof. Ng's videos (in the comments section to have clickable URLs).