DECISION TREES AND NEURAL NETWORKS FOR REASONING AND KNOWLEDGE ACQUISITION FOR AUTONOMOUS AGENTS

This paper addresses the problem of developing the domain knowledge ba se for autonomous manufacturing agents (subsystems). In an organisatio nal and behavioural context, autonomous agents consist of elements (pe ople, machines, robots, etc.) tied by the flow of information between an agent and its external environment as well as within an agent. The paper focuses on non-quantitative support that can be used for reasoni ng and retrieval of knowledge describing such a flow of information in various decision situations. The formal quantitative model can be use d to generate some examples of the above domain knowledge. Quantitativ e models of an information flow evaluation, however, are often too com plex to serve as the tools useful in the knowledge retrieval process. In the paper, the application of such artificial intelligence tools as decision trees and neural networks is investigated and illustrated wi th examples.