论文信息 - An Improved Training Algorithm for the Linear Ranking Support Vector Machine

An Improved Training Algorithm for the Linear Ranking Support Vector Machine

We introduce an O(ms+mlog(m)) time complexity method for training the linear ranking support vector machine, where m is the number of training examples, and s the average number of non-zero features per example. The method generalizes the fastest previously known approach, which achieves the same efficiency only in restricted special cases. The excellent scalability of the proposed method is demonstrated experimentally.

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

[2] Rudolf Bayer,et al. Symmetric binary B-Trees: Data structure and maintenance algorithms , 1972, Acta Informatica.

[3] Ronald L. Rivest,et al. Introduction to Algorithms , 1990 .

[4] Eyke Hllermeier,et al. Preference Learning , 2010 .

[5] Thorsten Joachims,et al. Training linear SVMs in linear time , 2006, KDD '06.

[6] Alexander J. Smola,et al. Bundle Methods for Regularized Risk Minimization , 2010, J. Mach. Learn. Res..

[7] Tapio Salakoski,et al. Training linear ranking SVMs in linearithmic time using red-black trees , 2010, Pattern Recognit. Lett..

[8] Clifford Stein,et al. Introduction to Algorithms, 2nd edition. , 2001 .

[9] Yiming Yang,et al. RCV1: A New Benchmark Collection for Text Categorization Research , 2004, J. Mach. Learn. Res..

[10] Alexander J. Smola,et al. Bundle Methods for Machine Learning , 2007, NIPS.

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

[12] Klaus Obermayer,et al. Support vector learning for ordinal regression , 1999 .

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

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