DECISION-MAKING WITH GREEDY MULTIAGENT CLASSIFIER

Decision support systems (DSS) which deal with the noisy data is in the scope of this paper. Types and sources of noise, the influence of noise on a decision-making process are considered. Traditional classification methods and their drawbacks are examined. Two main approaches to mitigate the effects produced by noise are presented and observed. A multiagent-based classifier based on a «greedy» strategy has been proposed. At the current stage of the research the generic methods of agent negotiation and collaboration are presented. Some typical situations of classification are considered and selected simulation results are given.