January 1, 2005

A support vector method for multivariate performance measures

Key Points

Key points are not available for this paper at this time.

Abstract

This paper presents a Support Vector Method for optimizing multivariate nonlinear performance measures like the F1-score. Taking a multivariate prediction approach, we give an algorithm with which such multivariate SVMs can be trained in polynomial time for large classes of potentially non-linear performance measures, in particular ROCArea and all measures that can be computed from the contingency table. The conventional classification SVM arises as a special case of our method.

Bookmark

Cite This Study

Thorsten Joachims (Sat,) studied this question.

synapsesocial.com/papers/6a0a4eb1df43cb70ca573f2a https://doi.org/https://doi.org/10.1145/1102351.1102399

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

1Statistical Learning Theory1999 · 26,957 citations
2Optimizing classifier performance via an approximation to the Wilcoxon-Mann-Whitney statistic2003 · 211 citations
3Probabilistic Outputs for Support vector Machines and Comparisons to Regularized Likelihood Methods1999 · 4,875 citations
4Max-Margin Markov Networks2003 · 1,251 citations
5Estimating Class Membership Probabilities using Classifier Learners.2005 · 36 citations

Bookmark