A Medical Diagnosis Method Based on Interval-valued Fuzzy Cognitive Map

Abstract

Cognitive map is a powerful and useful tool for medical diagnosis. However, traditional fuzzy cognitive map cannot comprehensively represent experts ideas and some significant information is lost during the process of defuzzification. To overcome these drawbacks, a novel model called the interval-valued fuzzy cognitive map is introduced. In the proposed model, interval-valued fuzzy sets, rather than fuzzy sets, are employed to represent the concept nodes with their weights. A numerical example of breast cancer risk prediction is provided to illustrate the validity of the proposed model. Results show that the proposed model can enhance the diagnostic accuracy to 92.5%.