Improvements on ν-Twin Support Vector Machine. (July 2016)

• Record Type: Journal Article
• Authors: Khemchandani, Reshma; Saigal, Pooja; Chandra, Suresh
• Is Part Of: Neural networks. Volume 79(2016)
• Page Start: 97
• Page End: 107
• Publication Date: 2016-07
• Subjects: Twin support vector machines -- Machine learning -- Unconstrained optimization -- Image pixel classification; Neural computers -- Periodicals; Neural networks (Computer science) -- Periodicals; Neural networks (Neurobiology) -- Periodicals; Nervous System -- Periodicals; Ordinateurs neuronaux -- Périodiques; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux neuronaux (Neurobiologie) -- Périodiques; Neural computers; Neural networks (Computer science); Neural networks (Neurobiology); Periodicals; 006.32
• DOI: 10.1016/j.neunet.2016.03.011
• Languages: English
• ISSNs: 0893-6080

Abstract

Abstract: In this paper, we propose two novel binary classifiers termed as "Improvements on ν -Twin Support Vector Machine: I ν -TWSVM and I ν -TWSVM (Fast)" that are motivated by ν -Twin Support Vector Machine ( ν -TWSVM). Similar to ν -TWSVM, I ν -TWSVM determines two nonparallel hyperplanes such that they are closer to their respective classes and are at least ρ distance away from the other class. The significant advantage of I ν -TWSVM over ν -TWSVM is that I ν -TWSVM solves one smaller-sized Quadratic Programming Problem (QPP) and one Unconstrained Minimization Problem (UMP); as compared to solving two related QPPs in ν -TWSVM. Further, I ν -TWSVM (Fast) avoids solving a smaller sized QPP and transforms it as a unimodal function, which can be solved using line search methods and similar to I ν -TWSVM, the other problem is solved as a UMP. Due to their novel formulation, the proposed classifiers are faster than ν -TWSVM and have comparable generalization ability. I ν -TWSVM also implements structural risk minimization (SRM) principle by introducing a regularization term, along with minimizing the empirical risk. The other properties of I ν -TWSVM, related to support vectors (SVs), are similar to that of ν -TWSVM. To test the efficacy of the proposed method, experiments have been conducted on a wide range of UCI and a skewed variation of NDC datasets. We have also given the application of I ν -TWSVM as a binary classifier for pixel classification of color images.