报告题目：Enhanced Semi-Linear Credibility Formula
As one of the earliest modern important application of Bayesian statistics, credibility theory was developed originally to calibrate the risk premium by optimally combining the individual risk experience with that of the whole population. Bühlmann and Gisler (2006) proposed a semi-linear credibility formula with truncation provided that the second moments exist. The optimal truncation point is determined by minimizing the square loss function. However, the explicit formula of the optimal truncation point was not given in their work, and the underlying philosophy behind this method is not clearly explained. In this talk, we close this gap by presenting a tail index to characterize the whole population in the group, and underpinning the explicit formula for the optimal truncation point asymptotically under quadratic loss criterion. Numerical examples are presented to illustrate our model-free truncation methods. Our method not only performs very well for small index, but also competitively compared with the simple Bühlmann model for large index. This is a joint work with Ka Chun Cheung and Philip Yam.
张艺赢现为香港大学统计与精算学系在读博士，于兰州大学数学与统计学院分别获得学士学位（2012年）和硕士学位（2015年）。目前的研究兴趣主要包括：最优再保险、信度理论、相依风险模型、随机比较和可靠性理论。已发表学术论文20余篇，其中在精算学领域顶级期刊 Insurance: Mathematics and Economics，Scandinavian Actuarial Journal, ASTIN Bulletin发表3篇，运筹学领域顶级期刊European Journal of Operations Research发表2篇。