Classification by pairwise coupling

作者：Trevor Hastie, Robert John Tibshirani · 发表于：The Annals of Statistics · 年份：1998 · DOI：10.1214/aos/1028144844 · 被引用次数：1301 · 研究领域：Advanced Statistical Methods and Models、Spectroscopy and Chemometric Analyses、Statistical Methods and Inference

We discuss a strategy for polychotomous classification that involves coupling the estimating class probabilities for each pair of classes, and estimates together. The coupling model is similar to the Bradley-Terry method for paired comparisons. We study the nature of the class probability estimates that arise, and examine the performance of the procedure in real and simulated data sets. Classifiers used include linear discriminants, nearest neighbors, adaptive nonlinear methods and the support vector machine.