Good Freely Available Videos on Machine Learning

Stanford University CS 229: Machine Learning presented by Andrew Ng. Course provides a broad introduction to machine learning and statistical pattern recognition.

"Topics include: supervised learning (generative/discriminative learning, parametric/non-parametric learning, neural networks, support vector machines); unsupervised learning (clustering, dimensionality reduction, kernel methods); learning theory (bias/variance tradeoffs; VC theory; large margins); reinforcement learning and adaptive control."

Hi,

How do you like these: http://videolectures.net/Top/#t=vl&l=en&q=machine%20learning ?

Best regards, Jochen

This set of lectures given at Google a few years ago is a good introduction, and includes an introduction to using R for data mining: Statistical Aspects of Data Mining (Stats 202) (from Stanford).

Download MIT videos from links at "Lecture (ISO)" section.

MIT 6.867 Machine Learning Fall 2005 Video Lectures. Instructors: Leslie P Kaelbling, Lee Wee Sun

http://smasvr.nus.edu.sg/browse.asp?b=e&d=SMA\2005-2006\sma5514-fall\

MIT 6.825 Artificial Intelligence Fall 2005. Prof. Lozano-Perez, Prof. Ng

http://smasvr.nus.edu.sg/browse.asp?b=e&d=SMA\2005-2006\sma5504-fall\

Machine Learning anda Data Mining (Freiburg University)

http://electures.informatik.uni-freiburg.de/portal/web/guest/detail/-/modulnavigation/view/75/3164/

Advanced AI Techniques (Freiburg University)

http://electures.informatik.uni-freiburg.de/portal/web/guest/detail/-/modulnavigation/view/1/349/

(Warning: referencing my own work here. Caveat emptor.)

"The Counter-Intuitive Properties of Ensembles for Machine Learning, or, Democracy Defeats Meritocracy" is a talk I gave at Google to summarize my personal lessons learned from a decade of doing research on that the topic. The abstract is:

Machine learning is the process of using past experience to predict the future. There are many machine learning methods; neural nets, support vector machines, decision trees. The design trade-offs in optimizing them is a tricky business, still more art than science.

"Ensembles" are a machine-learning meta-method that can be applied to most machine learning algorithms. Ensembles generally greatly improve accuracy, provably do no harm, reduce or remove most of the design issues, are admirably suited to parallel and distributed computation, and are delightfully weird and counter-intuitive.

This talk will provide a terse introduction to machine learning and then discuss the properties of ensembles; what they are, various theories on why they work, and how they can be simply applied to improve existing machine learning code in situ.

I think you already know videolectures.net. They have recorded the different installments of the machine learning summer school series (http://www.mlss.cc/) which contains many high-profile machine learning researchers giving introductory talks. Just search for "mlss".

• techtalks.tv has a bunch of machine learning stuff, including some recently recorded talks from ICML 2011
• While not precisely machine learning, these mathematics lectures from the MIT open courseware site can help you review or improve on various math topics.
• Khan Academy has a wide variety of courses that could also help improve your math skills

Andrew Ng gave a really good introductory talk recently. It is mostly about unsupervised connectionist learning in a wide variety of domains.

i know it has been posted, but I like to be a bit more specific.

This summer school was focused on Non Parametric methods, and there are great lectures.

I highly recommend Blei's on Topic Model and Teh on Non Parametric models (2nd lecture, he drifted a little on the first one)

http://videolectures.net/mlss09uk_cambridge/

This one in particular was good, dealing with graphical models:

http://videolectures.net/mlss06tw_roweis_mlpgm/

It's a shame he committed suicide.