A mathematical programming approach to SVM-based classification with label noise. (October 2022)

• Record Type: Journal Article
• Authors: Blanco, Víctor; Japón, Alberto; Puerto, Justo
• Is Part Of: Computers & industrial engineering. Volume 172:Part A(2022)
• Publication Date: 2022-10
• Subjects: Supervised classification -- SVM -- Mixed integer non linear programming -- Label noise; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108611
• Languages: English
• ISSNs: 0360-8352

Abstract

Abstract: In this paper we propose novel methodologies to optimally construct Support Vector Machine-based classifiers that take into account that label noise occur in the training sample. We propose different alternatives based on solving Mixed Integer Linear and Non Linear models by incorporating decisions on relabeling some of the observations in the training dataset. The first method incorporates relabeling directly in the SVM model while a second family of methods combines clustering with classification at the same time, giving rise to a model that applies simultaneously similarity measures and SVM. Extensive computational experiments are reported based on a battery of standard datasets taken from UCI Machine Learning repository, showing the effectiveness of the proposed approaches. Highlights: We present mathematical programming approaches to classify datasets with noisy labels. Two alternatives are proposed based on solving Mixed Integer Linear and Non Linear models. The proposals considers that observations can be relabeled simultaneously to find the SVM-classifier. Extensive computational experiments are reported showing the effectiveness of the approaches.