A Hybrid Fuzzy Regression Model for Optimal Loss Reserving in Insurance

Abstract

In this article, a Hybrid Fuzzy Regression Model with Asymmetric Triangular Fuzzy Coefficients and optimized $h-$value in Generalized Linear Models (GLM) framework have been developed. The weighted functions of Fuzzy Numbers rather than the Expected value of Fuzzy Number is used as a defuzzification procedure. We perform the new model on a numerical data (Taylor and Ashe, 1983) to predict incremental payments in loss reserving. We prove that the new Hybrid Model with the optimized $h-$value produce better results than the classical GLM according to the Reserve Prediction Error and Reserve Standard Deviation.