A non-parametric approach to extending generic binary classifiers for multi-classification. (October 2016)

• Record Type: Journal Article
• Authors: Santhanam, Venkataraman; Morariu, Vlad I.; Harwood, David; Davis, Larry S.
• Is Part Of: Pattern recognition. Volume 58(2016:Oct.)
• Page Start: 149
• Page End: 158
• Publication Date: 2016-10
• Subjects: Multi-classification -- Ensemble method -- One-vs-one -- Orthogonal subspace -- Non-parametric density estimation; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2016.04.008
• Languages: English
• ISSNs: 0031-3203

Abstract

Abstract: Ensemble methods, which combine generic binary classifier scores to generate a multi-classification output, are commonly used in state-of-the-art computer vision and pattern recognition systems that rely on multi-classification. In particular, we consider the one-vs-one decomposition of the multi-class problem, where binary classifier models are trained to discriminate every class pair. We describe a robust multi-classification pipeline, which at a high level involves projecting binary classifier scores into compact orthogonal subspaces, followed by a non-linear probabilistic multi-classification step, using Kernel Density Estimation (KDE). We compare our approach against state-of-the-art ensemble methods (DCS, DRCW) on 16 multi-class datasets. We also compare against the most commonly used ensemble methods (VOTE, NEST) on 6 real-world computer vision datasets. Finally, we measure the statistical significance of our approach using non-parametric tests. Experimental results show that our approach gives a statistically significant improvement in multi-classification performance over state-of-the-art. Abstract : Highlights: Ensemble methods combine binary classifiers to yield a multi-classification output. One-vs-one ensemble: binary classifiers trained to discriminate each class pair. We propose a robust non-parametric probabilistic one-vs-one ensemble method: KDEMRP. KDEMRP improves classification performance over state-of-the-art (DCS, DRCW). KDEMRP improvements are statistically significant.