关键词： Mlinex损失函数, 信度保费, 贝叶斯保费, 相合估计

Abstract:

In the classical credibility theory,symmetric loss function is usually treated as the criterion for evaluating the predicted result,while the degree of fitting is ignored.To make better application of credibility model,the paper solves the problem of high premium caused by symmetric loss function in classical credibility models via using Mlinex loss function,obtains the Bayes premium and the credibility premium based on Mlinex loss function,and checks the consistency of Bayes premium and credibility premium.Simulation shows that the premium based on Mlinex function is better than that based on the classical credibility model.

Key words: Mlinex loss function, credibility premium, Bayes premium, consistency estimate