New Vague Set Based Similarity Measure for Pattern Recognition

Abstract

By analyzing existing famous similarity measures, these similarity measures are neither reasonable nor capable of discriminating difference between vague sets in some cases. Therefore, the main purpose of this study was to propose a new similarity measure to solve the drawbacks of those similarity measures and apply it to handle pattern recognition problems, which are with imperfect and/or imprecise information. It has proved that the proposed similarity measure satisfies all properties in the axiomatic definition of similarity measure. Furthermore, it also has illustrated to compare and to justify that the proposed similarity measure performs the same or better than existing similarity measures between vague sets. Finally, an application of the proposed similarity measure has successfully demonstrated that the proposed similarity measure can overcome the drawback of the existing similarity measure for measuring degree of similarity between vague sets and is effective and efficient for applying in the context of recognizing patterns.