A multi-agent learning paradigm for medical data mining diagnostic workbench

Abstract

This chapter presents a model of practical data mining diagnostic workbench that intends to support real medical diagnosis by employing two cutting edge technologies - data mining and multi-agent. To fulfill this effort, i +DiaMAS - an intelligent and interactive diagnostic workbench with multi-agent strategy, has been designed and partially implemented, where multi-agent approach can perfectly compensate most data mining methods that are only capable of dealing with homogeneous and centralized data. i +DiaMAS provides an integrated environment that incorporates a variety of preprocessing agents as well as learning agents through interactive interface agent. The philosophy behind i +DiaMAS is to assist physicians in diagnosing new case objectively and reliably by providing practical diagnostic rules that acquired from heterogeneous and distributed historical medical data, where their new discoveries and correct diagnostic results can benefit other physicians. © 2009 Springer-Verlag US.