应数学系田波平教授的邀请，受国际合作处资助，印第安纳大学南本德分校关忠教授将于近日来访公司并做报告，欢迎感兴趣的师生参加。

报告主题1：Maximum Approximate Bernstein Likelihood Estimations of Densities Under a Density Ratio Model

报告1(1)：The approximate Bernstein polynomial model is used to estimate densities and ROC curve under a two-sample density ratio model. The EM algorithm to obtain maximum approximate Bernstein likelihood estimate is proposed and studied. Bootstrap method is proposed to construct confidence band for ROC curve. Simulation is conducted to compare the performance of the proposed method with the existing ones including the parametric method. The proposed method is illustrated by real data examples.

报告1(2)：Grouped continuous data are common in applications. Nonparametric smooth estimation of density and ROC curve is extremely difficult based on grouped continuous data. The approximate Bernstein polynomial model is a parametric solution of some nonparametric statistical problems and makes it possible to obtain smooth estimation of ROC curve.

报告1(3)：Further related topics. We can apply the change point method for estimating number of changes to any multiple change point model, parametric, semiparametric, and nonparametric.

报告主题2：变点理论及其应用浅谈

报告主题3：印第安纳大学硕士生和旗下产业浅谈

时间和地点：2018.07.01, 09:00-12:00，格物楼503，报告1(1)，1(2)

时间和地点：2018.07.01, 15:00-16:30，格物楼503，报告1(3)

时间和地点：2018.07.02, 14:00-17:00，格物楼522，报告主题2，3

报告人简介：关忠，印第安纳大学南本德分校终身教授，002全讯白菜网统计学兼职博导，美国托莱多大学统计学博士，耶鲁老员工物统计系博士后，美国统计协会会员，美国数学协会会员。曾任黑龙江省概率统计协会秘书长，中国应用统计协会黑龙江省分会理事。在统计学顶级期刊Biometrika上发表1篇文章（独立作者），在生物统计顶级期刊Biometrics上发表1篇文章，在Statistica Sinica, Journal of Nonparametric Statistics, Journal of Statistical Planning and Inference, Canadian Journal of Statistics等国际一流期刊上发表文章20余篇。