M. Arefi - Department of Statistics, Faculty of Mathematical Sciences and Statis- tics, University of Birjand, Birjand, Iran
S. M. Taheri - Department of Engineering Science, College of Engineering, Univer- sity of Tehran, Tehran, P.O. Box ۱۱۳۶۵-۴۵۶۳, Iran

A new approach to define the similarity measure betweeninterval-valued fuzzy sets is presented. The proposed approach isbased on a weighted measure in which the normalized similaritiesbetween lower functions and also between upper functions arecombined by a weight parameter. The properties of this similaritymeasure are investigated. It is shown that, the proposed measurehas some advantages in comparison with  the  commonly usedsimilarity measures.

Interval-valued fuzzy set, Intuitionistic fuzzy set, Pattern recognition, Similarity measure