U-Processes and Preference Learning

Preference learning has caused great attention in machining learning. In this letter we propose a learning framework for pairwise loss based on empirical risk minimization of U-processes via Rademacher complexity. We first establish a uniform version of Bernstein inequality of U-processes of degree 2 via the entropy methods. Then we estimate the bound of the excess risk by using the Bernstein inequality and peeling skills. Finally, we apply the excess risk bound to the pairwise preference and derive the convergence rates of pairwise preference learning algorithms with squared loss and indicator loss by using the empirical risk minimization with respect to U-processes.

[1]  M. A. Arcones,et al.  A Bernstein-type inequality for U-statistics and U-processes , 1995 .

[2]  E. Giné,et al.  Decoupling: From Dependence to Independence , 1998 .

[3]  Colin McDiarmid,et al.  Concentration for self-bounding functions and an inequality of Talagrand , 2006 .

[4]  Yoram Singer,et al.  An Efficient Boosting Algorithm for Combining Preferences by , 2013 .

[5]  A. W. van der Vaart,et al.  Uniform Central Limit Theorems , 2001 .

[6]  M. Talagrand.  New concentration inequalities in product spaces , 1996 .

[7]  E. Rio,et al.  Concentration around the mean for maxima of empirical processes , 2005, math/0506594.

[8]  Tong Zhang,et al.  Statistical Analysis of Bayes Optimal Subset Ranking , 2008, IEEE Transactions on Information Theory.

[9]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[10]  M. Talagrand,et al.  Probability in Banach spaces , 1991 .

[11]  Wojciech Rejchel,et al.  On Ranking and Generalization Bounds , 2012, J. Mach. Learn. Res..

[12]  P. Massart.  Some applications of concentration inequalities to statistics , 2000 .

[13]  P. Gänssler.  Weak Convergence and Empirical Processes - A. W. van der Vaart; J. A. Wellner. , 1997 .

[14]  V. Koltchinskii.  Local Rademacher complexities and oracle inequalities in risk minimization , 2006, 0708.0083.

[15]  Peter L. Bartlett,et al.  Rademacher and Gaussian Complexities: Risk Bounds and Structural Results , 2003, J. Mach. Learn. Res..

[16]  E. Giné,et al.  Limit Theorems for $U$-Processes , 1993 .

[17]  M. Talagrand.  Sharper Bounds for Gaussian and Empirical Processes , 1994 .

[18]  Yoram Singer,et al.  Learning to Order Things , 1997, NIPS.

[19]  V. Koltchinskii,et al.  Rademacher Processes and Bounding the Risk of Function Learning , 2004, math/0405338.

[20]  Tie-Yan Liu,et al.  Learning to Rank for Information Retrieval , 2011 .

[21]  Colin Campbell,et al.  Rademacher Chaos Complexities for Learning the Kernel Problem , 2010, Neural Computation.

[22]  P. Massart,et al.  About the constants in Talagrand's concentration inequalities for empirical processes , 2000 .

[23]  V. Koltchinskii.  Rejoinder: Local Rademacher complexities and oracle inequalities in risk minimization , 2006, 0708.0135.

[24]  S. Boucheron,et al.  A sharp concentration inequality with applications , 1999, Random Struct. Algorithms.

[25]  Eyke Hüllermeier,et al.  Label ranking by learning pairwise preferences , 2008, Artif. Intell..

[26]  V. Koltchinskii,et al.  Oracle inequalities in empirical risk minimization and sparse recovery problems , 2011 .

[27]  Shivani Agarwal,et al.  Stability and Generalization of Bipartite Ranking Algorithms , 2005, COLT.

[28]  T. Salakoski,et al.  Learning to Rank with Pairwise Regularized Least-Squares , 2007 .

[29]  D. Pollard,et al.  $U$-Processes: Rates of Convergence , 1987 .

[30]  O. Bousquet.  A Bennett concentration inequality and its application to suprema of empirical processes , 2002 .

[31]  G. Lugosi,et al.  Ranking and empirical minimization of U-statistics , 2006, math/0603123.

[32]  P. Bartlett,et al.  Local Rademacher complexities , 2005, math/0508275.

[33]  M. Ledoux.  On Talagrand's deviation inequalities for product measures , 1997 .

[34]  Jon A. Wellner,et al.  Weak Convergence and Empirical Processes: With Applications to Statistics , 1996 .