Intuitionistic Fuzzy Universum Support Vector Machine

Abstract

The classical support vector machine is an effective classification technique. It solves a convex optimization problem to give a global solution. But it suffers from noise and outliers. To deal with this, an intuitionistic fuzzy number (IFN) is assigned to the training samples which reduces the effect of noise. In this paper, we propose intuitionistic fuzzy universum support vector machine (IFUSVM), where IFN is assigned to the training data points in presence of universum data. Universum points lead to prior knowledge about data distribution and assignment of IFN to the data points reduces the effect of outliers and noise. Thus, leading to the enhanced generalization property of the model. Numerical experimental results and statistical analysis over 17 binary benchmark UCI datasets show the superiority of the proposed model over the baseline models in terms of rank and accuracy. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.