Constructing n-cell numbers by using double-side separation degrees and pattern recognition based on the maximal membership principle

Abstract

In this study, by using both the left and right separation degrees, the authors establish a method constructing fuzzy n-cell numbers to express uncertain or imprecise multi-channel digital signals. They conclude that the maximal membership method can be adopted to identify an object in some precise or certain environment. Moreover, the authors point out that the method based on the maximal membership possesses some weaknesses, and introduce the concept of p-membership function to overcome the shortcomings, thus using the maximal p-membership principle to identify an object.