講座內(nèi)容： To effectively representdeterministic uncertainties and vagueness as well as human subjectiveobservation and judgment encountered in many real-world problems especiallythose in medicine, we recently originated a theory of fuzzy discrete event systems(DES). We introduced fuzzy states and fuzzy event transition and generalizedconventional crisp DES to fuzzy DES. The largely graph-based framework of thecrisp DES was unsuitable for the expansion and we thus reformulated it usingstate vectors and event transition matrices which could be extended to fuzzyvectors and matrices by allowing their elements to take values in [0, 1]. Wealso extended optimal control of DES to fuzzy DES. The new fuzzy DES theory isconsistent with the existing theory, both at conceptual and computation levels,in that the former contains the latter as a special case when the membershipgrades are either 0 or 1. We further developed the FDES theory so that itpossessed self-learning capability.

We have applied the fuzzyDES theory to develop an innovative software system for medical treatment,specifically for the first round of highly active antiretroviral therapy ofHIV/AIDS patients. The objective is to build such a system whose treatmentregimen choice for any given patient will match expert AIDS physician’sselection to produce the (anticipated) optimal treatment outcome. Preliminaryretrospective evaluation of our prototype system using patients treated in ourinstitution’s AIDS Clinical Center demonstrates encouraging results when thesystem operates in either self-learning mode or non-learning mode. Our approachhas the capabilities of generalizing, learning, representing knowledge even inthe face of weak consensus of domain experts, and being readily upgradeable to newmedical knowledge. These are practically important features to medicalapplications in general, and HIV/AIDS treatment in particular, as nationalHIV/AIDS treatment guidelines are modified several times per year.

This research wassupported in part by the National Institutes of Health under grant R21 EB001529-01A1, and by Wayne StateUniversity under a Research Enhancement Program grant.