Analogical gates - a network approach to multi-criteria aggregation with application in risk assessment
Analogical gates are developed to implement four-quadrant fuzzy-generalisation of Boolean logical gates. They provide an alternative to the traditional rule-based realisation of fuzzy inference. Therefore, analogical gates can be used in a wide range of applications especially for multi-criteria aggregation. In this paper, the basic properties of the analogical gates are reviewed, their advantages compared to the traditional fuzzy formalism are discussed and some preliminary results on synthesis are elaborated. To demonstrate their benefits, analogical gates are applied in the risk assessment of twenty industrial plants for insurance purposes. The results are compared to those made by human experts and are found to match their assessments with a good accuracy.
Keywords: multi-, criteria aggregation, risk assessment, analogical gates, insurance