摘要

In this paper, we present an evolutionary autonomous agent approach to associative classification (EAA-AC) which utilises evolutionary autonomous agents. In this approach, the distributed computational entities, autonomous agents, operate in the N-dimensional rules space and exhibit a number of reactive behaviours. To effectively locate the feature points, class association rules (CARs), individual agents sense the local stimuli from the environment by means of calculating the confidence value, and accordingly activate their behaviours. The behavioural repository of the agents consists of self-reproduction, directional diffusion, rule-marking and death. We evaluate the performance of the proposed approach for AC and the results have shown that it is efficient in dealing with the problem on the complexity of the rule search space.

  • 出版日期2011

全文