A new multi agent system based on online sequential extreme learning machines and Bayesian Formalism

Abstract

In this article, a new Multi Agent System based on Online Sequential Extreme Learning Machine (OSELM) neural network and Bayesian Formalism (MAS-OSELM-BF) is introduced. It is an improvement of a single OSELM (single agent) by combined multiple OSELMs (multi agents) with a final decision making module (parent agent). Here, the development of the parent agent is motivated by the Bayesian Formalism. A series of empirical studies to assess the effectiveness of the MAS-OSELM-BF in pattern classification tasks is conducted. The results demonstrated that the MAS-OSELM-BF able to produce good performance as compared with a single OSELM and other method that employed ensemble OSLEM (EOSELM). © 2011 IEEE.