Today, the machine learning class started teaching the theory of machine learning. So I want to collect some materials related to it here. And I think I will update this post along with the class.
This paper discusses a lot of theory behind classification, which is one main part of machine learning.
2, S. Boucheron, O. Bousquet, and G. Lugosi, (2004), Concentration inequalities. (PDF,POSTSCRIPT) in O. Bousquet, U.v. Luxburg, and G. Rätsch (editors), Advanced Lectures in Machine Learning, Springer, pp. 208–240, 2004.
This is the content of the first class on the theory of machine learning. And the author also has the following lecture notes on this topic:
Concentration-of-measure inequalities presented at the
Machine Learning Summer School 2003, Australian National University, Canberra,
at the Workshop on combinatorics, probability and algorithms 2003, held at the CRM in Montreal,
at the Winter School on Probabilistic Methods in High Dimension Phenomena , Toulouse, January 10-14, 2005
and at the Workshop de Combinatória e Concentração de Medida IMPA, Rio de Janeiro, February 23-25, 2005 (PDF,POSTSCRIPT).
3, O. Bousquet, S. Boucheron, and G. Lugosi, (2004), Introduction to statistical learning theory. (PDF,POSTSCRIPT) in O. Bousquet, U.v. Luxburg, and G. Rätsch (editors), Advanced Lectures in Machine Learning, Springer, pp. 169–207, 2004.
4, S. Kulkarni, G. Lugosi, and S. Venkatesh (1998), Learning Pattern Classification—A Survey. (POSTSCRIPT) 1948–1998 Special Commemorative Issue of IEEE Transactions on Information Theory. , vol.44, 2178–2206. Reprinted in S. Verdú, S.W. McLaughlin (editors.), Information Theory: 50 Years of Discovery, IEEE Press, New York, 1999.
So far these materials all come from one professor GÁBOR LUGOSI.
For concentration inequalities, there is also a chapter from one book: Old and new concentration inequalities, whose author is Fan Chung Graham. In addition, there is a survey about this: Concentration inequalities and martingale inequalities — a survey .
Videolectures on concentration inequalities with machine learning applications
There is a course: STAT 598Y: Statistical Learning Theory given in Statistics Department at Purdue University. It focuses more on the machine learning theory.
Foundations of Statistical Learning Theory : Empirical Infe-rence in high-dimention spaces (videolecture)
the slides of statstical learning theory given by Olivier Bousquet
Peter L. Bartlett, Shahar Mendelson; 3(Nov):463-482, 2002.
Rademacher Complexity—Marcel Lüthi, Thomas Vetter
Algebraic Geometry and Statistical Learning Theory (Cambridge Monographs on Applied and Computational Mathematics)
|by: Martin Anthony, Peter L. Bartlett|