A Novel Approach to Parameter Reduction of Fuzzy Soft Set

The fuzzy soft set (FSS) that combines soft set theory with fuzzy set theory has been introduced to deal with uncertainty in many practical decision-making problems. However, there exist some less important and superfluous information in the decision-making process which should be removed. In order to settle these problems, different methods have been developed for parameter reduction of FSS which are based on different decision criteria. In this paper, the problem of parameter reduction of FSS is studied based on choice value criteria. Initially, some previous reduction approaches such as; normal parameter reduction (NPR), proximate normal parameter reduction (PNPR) and distance-based parameter reduction (DBPR) of FSS are analyzed and their inherent difficulties are discussed. Then, a new method for parameter reduction of FSS is proposed and its heuristic algorithm is given. The proposed algorithm is compared with NPR, PNPR and DBPR algorithms from the aspects of success rate of finding reduction, reduction ratio and decision ability, respectively. The comparison results show that the proposed algorithm has much higher applicability and efficiency as compared to rest of the three algorithms. Finally, an application of our proposed algorithm is discussed in a real life decision-making problem.