An Improved Extreme Learning Machine Based on Full Rank Cholesky Factorization

¹ School of Mathematics & Statistics, Guizhou University of Finance and Economics, Guiyang, Guizhou, China
² School of Artificial Intelligence, Xidian University, Xi’an, Shaanxi, China
³ Department of Mathematics and Physics, Changzhou Campus, Hohai University, Changzhou, Jiangsu, China

^a Corresponding author: liuzuo.zhi@163.com

Abstract

Extreme learning machine (ELM) is a new novel learning algorithm for generalized single-hidden layer feedforward networks (SLFNs). Although it shows fast learning speed in many areas, there is still room for improvement in computational cost. To address this issue, this paper proposes an improved ELM (FRCFELM) which employs the full rank Cholesky factorization to compute output weights instead of traditional SVD. In addition, this paper proves in theory that the proposed FRCF-ELM has lower computational complexity. Experimental results over some benchmark applications indicate that the proposed FRCF-ELM learns faster than original ELM algorithm while preserving good generalization performance.