Abductive design for a mixture of two components

In Tanaka and Naka (1996), a methodology was presented to overcome the difficulty in predicting nonlinear cooperative effects due to mixing components. The basic framework was adapted to consecutive human knowledge representation and acquisition, integrating two methodologies, such as fuzzy clustering as pattern recognition (knowledge representation) and logical procedure (knowledge acquisition). In this paper, three inference patterns of deduction, induction and abduction, are developed in accordance with Peirce's terminology.