Building a hybrid method for analyzing the risk by integrating the fuzzy logic and the improved fuzzy clustering algorithm FCM-R

Abstract

Analysis of risk is a problem has attracted a special deal of consideration from enterprises. This paper proposes the integrating the fuzzy logic and the improved fuzzy clustering algorithm FCM-R to build a hybrid method for analyzing the risk. The method has two stages. In stage 1, first, the fuzzy logic is used to determine the risk levels of objects. The next, the objects which have risk level is greater or equal a given threshold, are selected for analyzing in the next stage. In stage 2, the improved fuzzy clustering algorithm FCM-R is applied to the objects chosen to create the appropriate number of clusters which are ranked based on the risk level measure of clusters. Here, for illustrating, we choose objects to analyze risk are customers in the enterprise. The proposed method has been experimented with real data set to generate customer clusters ranked according to the risk level measure from high to low. The results will be to use for predicting the customer’s risk and will be to help for offering the risk management policies to avoid loss.