Flow Graphs and Decision Tables with Fuzzy Attributes

This paper is concerned with the issue of design and analysis of fuzzy decision systems, basing on recorded process data. A concept of fuzzy flow graphs is proposed to allow representation of decision tables with fuzzy attributes. Basic notions of the crisp flow graph approach are generalized. Satisfaction of flow graph properties, with respect to fuzzy connectives used in calculations, is taken into account. Alternative definitions of the path's certainty and strength are introduced. In an illustrative example a decision table with fuzzy attributes is analyzed and interpreted in terms of flow graphs.