Robust Least Squares Support Vector Machine

• 발행기관: 한국자료분석학회
• 저자명: Daehak Kim
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.6 No.4, 943~951쪽, 전체 9쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2004.08.30

영문 초록

The least squares support vector machine(LS-SVM) is a widely applicable and useful machine learning technique for classification and regression analysis. LS-SVM is an SVM(support vector machine) version which involves equality constraints instead of inequality constraints and works with a squared loss function, which leads the solution to be obtained from a linear Karush-Kuhn-Tucker conditions instead of a quadratic programming problem. A drawback of LS-SVM is that the estimates is less robust due to the assumption of the errors and the use of a squared loss function. In this paper we propose a robust LS-SVM which imposes the robustness on the estimation of LS-SVM by eliminating the candidates of outliers. The proposed method are also applied to pruning support vectors in the LS-SVM case. In the numerical studies, the performance of the robust LS-SVM is shown and compared with the ordinary LS-SVM via FVU.

목차

1. Introduction

2. Robust LS-SVM Regression

3. Robust Pruning Procedure

4. Numerical Study

5. Concluding Remarks

References

키워드

SVM LS-SVM Karush-Kuhn-Tucker conditions Outlier Robust estimation contaminated distribution Fraction of Variance Unexplained