Active Set Fuzzy Support Vector ϵ-Insensitive Regression Approach

Abstract

In this paper a new fuzzy linear support vector machine formulation for regression problems is proposed and solved by the active set computational strategy. In this model, to each input data a fuzzy membership value is associated so that the input data can contribute proportionally to the learning of the decision surface. The proposed method has the advantage that its solution is obtained by solving a system of linear equations at a finite number of times rather than solving a quadratic optimization problem. Numerical experiments have been performed and the results obtained are in close agreement with the exact solution of the problems considered which clearly shows the effectiveness of the method.