A novel ELM K-means algorithm for clustering

Abstract

Extreme learning machine (ELM) as a new technology has shown its good generalization performance in regression and classification applications. Clustering analysis is an important tool to explore the structure of data and has been employed in many disciplines and applications. In this work, we propose a method that efficiently performs clustering in a high-dimensional space. The method builds on ELM projection into a high-dimensional feature space and the K-means algorithm for unsupervised clustering. The proposed ELM K-means algorithm is tested on twelve benchmark data sets. The experimental results indicate that ELM K-means algorithm can efficiently be used for multivariate data clustering.