A combined SVM and LDA approach for classification

TLDR

The SVM/LDA classifier extends support vector machines by incorporating global data information and generalizes linear discriminant analysis by adding local margin maximization. The paper introduces the SVM/LDA classifier, a new large‑margin model. The authors implement SVM/LDA by leveraging existing SVM software and evaluate it against SVM and LDA on synthetic and real‑world benchmarks. Empirical results demonstrate that SVM/LDA can be solved with standard SVM tools and performs comparably to or better than SVM and LDA on benchmark datasets.

Abstract

This paper describes a new large margin classifier, named SVM/LDA. This classifier can be viewed as an extension of support vector machine (SVM) by incorporating some global information about the data. The SVM/LDA classifier can be also seen as a generalization of linear discriminant analysis (LDA) by incorporating the idea of (local) margin maximization into standard LDA formulation. We show that existing SVM software can be used to solve the SVM/LDA formulation. We also present empirical comparisons of the proposed algorithm with SVM and LDA using both synthetic and real world benchmark data.