报告题目：Modelling losses using mixture of Log-normal and Inverse Gaussian distribution

报告人：Vesna Rajić， 教授 贝尔格莱德大学

邀请人：李伟

报告时间：2021年4月29日15:50-16:35

腾讯会议 ID：951 648 454

报告人简介：Vesna Rajić，塞尔维亚贝尔格莱德大学经济学院的年轻教授，1997年本科毕业于贝尔格莱德大学数学学院理论数学专业， 2002年在贝尔格莱德大学数学学院获得概率统计专业硕士学位，2007年获得贝尔格莱德大学经济学院统计科学专业博士学位，并留校讲授精算数学与统计分析方面的课程，具有扎实的数学理论基础和丰富的统计教学经验。她的科研课题主要集中在应用数学与统计、非线性分析、精算数学方面。目前担任塞尔维亚本国期刊“Ekonomika preduzeća”以及国际著名期刊“Journal of Statistics: Advances in Theory and Applications“，”Journal of Economic and Social Studies”的编委会委员，著有Risk measurement and control in insurance以及Quantitative Models in Economics两本专著。是塞尔维亚统计学会会员；塞尔维亚数学学会会员；经济学院理事会理事；经济学家科学学会会员；教授委员会委员，也是7个会议的项目委员会成员。是Neural Computing and Applications; FPTA; Journal of Applied Mathematics; Journal of Uncertainty Analysis and Applications; Journal of Statistical Computation and Simulation; Journal of Applied Statistics; Yujor; Economic Annals; Ekonomika preduzeća这些杂志的审稿人。

报告摘要：Obtaining accurate estimates of the losses that could arise from insurance contracts is of outstanding importance for insurance companies. On the basis of available loss data derived probability distributions determine insurance premiums, reserves, solvency capital and company’s retention. The use of probabilistic models which describe loss data to the best possible extent is a precondition for sound and safe operations of insurers. Here, we suggest the use of mixture of Inverse Gaussian and Log-normal distribution to model insurance losses. Expectation-maximization algorithm is used for estimating the unknown parameters. Goodness-of-fit tests are performed for industrial fire insurance loss data. Formulas for Value-at-Risk and conditional tail expectation calculation are provided for individual and aggregate losses.