Wu, Ting-Fan; Lin, Chih-Jen; Weng, Ruby C. Probability estimates for multi-class classification by pairwise coupling. (English) Zbl 1222.68336 J. Mach. Learn. Res. 5, 975-1005 (2004). Summary: Pairwise coupling is a popular multi-class classification method that combines all comparisons for each pair of classes. This paper presents two approaches for obtaining class probabilities. Both methods can be reduced to linear systems and are easy to implement. We show conceptually and experimentally that the proposed approaches are more stable than the two existing popular methods: voting and the method by T. Hastie and R. Tibshirani [Ann. Stat. 26, No. 2, 451–471 (1998; Zbl 0932.62071)]. Cited in 47 Documents MSC: 68T05 Learning and adaptive systems in artificial intelligence 62H30 Classification and discrimination; cluster analysis (statistical aspects) Keywords:pairwise coupling; probability estimates; random forest; support vector machines Citations:Zbl 0932.62071 Software:LIBSVM; MNIST PDFBibTeX XMLCite \textit{T.-F. Wu} et al., J. Mach. Learn. Res. 5, 975--1005 (2004; Zbl 1222.68336) Full Text: Link