Séparateurs à Vaste Marge pondérés en norme L2 pour la sélection de variables en apprentissage d'ordonnancement

Les algorithmes d’apprentissage d’ordonnancement utilisent un tres grand nombre de caracteristiques pour apprendre les fonctions d’ordonnancement, entrainant une augmentation des temps d’execution et du nombre de caracteristiques redondantes ou bruitees. La selection de variables est une methode prometteuse pour resoudre ces enjeux. Dans cet article, nous pro- posons de nouvelles methodes de selection de variables en apprentissage d’ordonnancement basees sur des approches de ponderation des SVM en norme l2. Nous proposons une adap- tation d’une methode l2-AROM pour la resolution des SVM en norme l0 et un algorithme generique de ponderation de la norme l2 qui resout les problemes en norme l0 et l1. Nos ex- perimentations montrent que les methodes proposees sont jusqu’a 7 fois plus rapides et 10 fois plus parcimonieuses que l’etat de l’art, pour des qualites d’ordonnancement equivalentes.

[1]  Tie-Yan Liu,et al.  Learning to rank: from pairwise approach to listwise approach , 2007, ICML '07.

[2]  Yong Tang,et al.  FSMRank: Feature Selection Algorithm for Learning to Rank , 2013, IEEE Transactions on Neural Networks and Learning Systems.

[3]  S. Sathiya Keerthi,et al.  Efficient algorithms for ranking with SVMs , 2010, Information Retrieval.

[4]  Thorsten Joachims,et al.  Optimizing search engines using clickthrough data , 2002, KDD.

[5]  Eric R. Ziegel,et al.  The Elements of Statistical Learning , 2003, Technometrics.

[6]  Wook-Shin Han,et al.  Efficient feature weighting methods for ranking , 2009, CIKM.

[7]  Jie Wu,et al.  Sparse Learning-to-Rank via an Efficient Primal-Dual Algorithm , 2013, IEEE Transactions on Computers.

[8]  Tao Qin,et al.  Feature selection for ranking , 2007, SIGIR.

[9]  Tapio Pahikkala,et al.  An efficient algorithm for learning to rank from preference graphs , 2009, Machine Learning.

[10]  T. Pahikkala.  Greedy RankRLS : a Linear Time Algorithm for Learning Sparse Ranking Models , 2010 .

[11]  Larry A. Rendell,et al.  A Practical Approach to Feature Selection , 1992, ML.

[12]  Yiqun Liu,et al.  Hierarchical feature selection for ranking , 2010, WWW '10.

[13]  Hang Li,et al.  AdaRank: a boosting algorithm for information retrieval , 2007, SIGIR.

[14]  Wotao Yin,et al.  Iteratively reweighted algorithms for compressive sensing , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[15]  Bernhard Schölkopf,et al.  Use of the Zero-Norm with Linear Models and Kernel Methods , 2003, J. Mach. Learn. Res..

[16]  Josiane Mothe,et al.  Nonconvex Regularizations for Feature Selection in Ranking With Sparse SVM , 2014, IEEE Transactions on Neural Networks and Learning Systems.

[17]  Chih-Jen Lin,et al.  LIBLINEAR: A Library for Large Linear Classification , 2008, J. Mach. Learn. Res..

[18]  Yong Tang,et al.  Greedy feature selection for ranking , 2011, Proceedings of the 2011 15th International Conference on Computer Supported Cooperative Work in Design (CSCWD).

[19]  Gregory N. Hullender,et al.  Learning to rank using gradient descent , 2005, ICML.

[20]  Feng Pan,et al.  Feature selection for ranking using boosted trees , 2009, CIKM.

[21]  Ron Kohavi,et al.  Wrappers for Feature Subset Selection , 1997, Artif. Intell..

[22]  Nick Kingsbury,et al.  FAST L0-based sparse signal recovery , 2010, 2010 IEEE International Workshop on Machine Learning for Signal Processing.

[23]  Denyse Baillargeon,et al.  Bibliographie , 1929 .

[24]  W. Bruce Croft,et al.  Feature Selection for Document Ranking using Best First Search and Coordinate Ascent , 2010 .

[25]  Tapio Elomaa,et al.  A Walk from 2-Norm SVM to 1-Norm SVM , 2009, 2009 Ninth IEEE International Conference on Data Mining.

[26]  Tao Qin,et al.  Robust sparse rank learning for non-smooth ranking measures , 2009, SIGIR.

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