Sparse least-squares Universum twin bounded support vector machine with adaptive Lp-norms and feature selection

Hossein Moosaei, Fatemeh Bazikar, Milan Hladík, Panos M. Pardalos

Abstract: Highlights•Proposing a method to perform feature selection and classification simultaneously.•Proving some theorems for counting and finding the number of selected features.•Solving only systems of linear equations to find the optimal solution.•Extensive experimental results demonstrate the efficiency of the proposed method.