Abstract
A similarity measure is used as one of the approaches in comparing fuzzy numbers. Various similarity measures with distance functions and set theoretic elements have been introduced in the literature. However, most of the similarity measures are unable to discriminate two fuzzy numbers with respect to their location. A new similarity measure for comparing generalized trapezoidal fuzzy numbers is proposed. Four features are included in the similarity measure, namely the geometric distance, the center of gravity, Hausdorff distance and Dice similarity coefficient. The properties of the similarity measure are studied. A comparative analysis between the new similarity measure and some existing similarity measures are made over 16 sets of pairs of fuzzy numbers. The analysis shows that the proposed similarity measure is comparable to the existing methods. Furthermore, it has an advantage of able to discriminate the similarity of fuzzy numbers with different horizontal position which other similarity measure methods fail to perform.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call