Training linear ranking SVMs in linearithmic time using red–black trees

We introduce an efficient method for training the linear ranking support vector machine. The method combines cutting plane optimization with red–black tree based approach to subgradient calculations, and has O(ms + mlog (m)) time complexity, where m is the number of training examples, and s the average number of non-zero features per example. Best previously known training algorithms achieve the same efficiency only for restricted special cases, whereas the proposed approach allows any real valued utility scores in the training data. Experiments demonstrate the superior scalability of the proposed approach, when compared to the fastest existing RankSVM implementations.

► We introduce a fast training algorithm for the linear RankSVM.
► The algorithm is based on cutting plane optimization and binary search trees.
► The algorithm has linearithmic scaling, with no restrictions on utility scores.
► Experiments demonstrate orders of magnitude improvements in training time.