Dear Experts,

I have a query regarding the conversion of crisp value to fuzzy number.

My idea: Actually, I am deriving the weight of criteria using Shannon Entropy and the weight obtained through SE is converted to fuzzy numbering system, either triangular and trapezoidal membership functions based on data characteristics.

I am just wondering whether this approach is scientifically correct or not. Actually, there are some literatures where they used Fuzzy number within the AHP itself and also Fuzzy Entropy but in my case its a different case where the entropy weight is being converted to fuzzy number and then defuzzify to obtained the final crisp value back.

Please comments with your suggestions and recommendation.

Thanks in advance