Quantification of Perception Clusters Using R-Fuzzy Sets and Grey Analysis

This paper investigates the use of the R-fuzzy significance measure hybrid approach introduced by the authors in a previous work; used in conjunction with grey analysis to allow for further inferencing, providing a higher dimension of accuracy and understanding. As a single observation can have a multitude of different perspectives, choosing a single fuzzy value as a representative becomes problematic. The fundamental concept of an R-fuzzy set is that it allows for the collective perception of a populous, and also individualised perspectives to be encapsulated within its membership set. The introduction of the significance measure allowed for the quantification of any membership value contained within any generated R-fuzzy set. Such is the pairing of the significance measure and the R-fuzzy concept, it replicates in part, the higher order of complex uncertainty which can be garnered using a type-2 fuzzy approach, with the computational ease and objectiveness of a typical type-1 fuzzy set. This paper utilises the use of grey analysis, in particular, the use of the absolute degree of grey incidence for the inspection of the sequence generated when using the significance measure, when quantifying the degree of significance fore each contained fuzzy membership value. Using the absolute degree of grey incidence provides a means to measure the metric spaces between sequences. As the worked example will show, if the data contains perceptions from clusters of cohorts, these clusters can be compared and contrasted to allow for a more detailed understanding of the abstract concepts being modelled.