Approximate Reasoning in MAS: Rough Set Approach

In modeling multiagent systems for real-life problems, techniques for approximate reasoning about vague concepts and dependencies (ARVCD) are necessary. We discuss an approach to approximate reasoning based on rough sets. In particular, we present a number of basic concepts such as approximation spaces, concept approximation, rough inclu-sion, construction of information granules in calculi of in-formation granules, and perception logic. The approach to ARVCD is illustrated by examples relative to interactions of agents, ontology approximation, adaptive hierarchical learning of compound concepts and skills, behavioral pat-tern identification, planning, conflict analysis and negotia-tions, and perception-based reasoning. The concept approximation problem is the basic prob-lem investigated in machine learning, pattern recognitio