大数据中心系列报告workshop

发布时间:2016-11-21浏览次数:49

以上海财经大学99周年校庆学术月活动为契机,为增强学术交流,学院特邀请了国内资深统计学专家来我院访问,并开展大数据中心系列报告workshop。欢迎各位老师和同学参加!


时间:2016年11月9日 13:30-16:30

地点: 统计与管理学院1208会议室

活动安排:

日期

时间

地点

讲座专家

所在单位

讲座题目

11-9

13:30-14:30

统管学院1208

陈  敏

中国科学院

Functional Partial Linear Single-index Model

11-9

14:30-15:30

统管学院1208

李  元

广州大学

Study On double autoregressive moving average models

11-9

15:30-16:30

统管学院1208

孙六全

中国科学院

Mark-specific additive hazards regression with continuous marks



报告摘要(按报告人姓氏拼音排序)

Functional Partial Linear Single-index Model

陈敏中国科学院

This paper deals with the problem of predicting the real-valued response variableusing explanatory variables containing both multivariate random variable and random curve. Theproposed functional partial linear single-index model treats the multivariate random variable aslinear part and the random curve as functional single-index part, respectively. To estimate thenon-parametric link function, the functional single-index and the parameters in the linear part, atwo-stage estimation procedure is proposed. Compared with existing semi-parametric methods, theproposed approach requires no initial estimation and iteration. Asymptotical properties are establishedfor both the parameters in the linear part and the functional single-index. The convergencerate for the non-parametric link function is also given. In addition, asymptotical normality of theerror variance is obtained that facilitates the construction of confidence region and hypothesis testingfor the unknown parameter. Numerical experiments including simulation studies and a real-dataanalysis are conducted to evaluate the empirical performance of the proposed method.


Study On double autoregressive moving average models

李元,广州大学

Moving average models and vector autoregressive models with GARCH errors are studied respectively. Sufficient conditions of ergodicity are obtained for both models. Parameters of models are estimated by the quasi maximum  likelihood method. It is shown that estimators are asymptotically normal without assumption of existence of second moment conditions for observed time series.


Mark-specific additive hazards regression with continuous marks

孙六全,中国科学院

For survival data, mark variables are only observed at uncensored failure times,and it is of interest to investigate whether there is any relationship between the failure time and the mark variable. The additive hazards model, focusing on hazard differences ratherthan hazard ratios, has been widely used in practice. In this article,we propose a mark-specific additive hazards modelin which both the regression coefficient functions and the baseline hazard functiondepend nonparametrically on a continuous mark.An estimating equation approach is developed to estimate the regression functions,and the asymptotic properties of the resulting estimators are established.In addition, some formal hypothesis tests are constructed forvarious hypotheses concerning the mark-specific treatment effects.The finite sample behavior of the proposed estimators is evaluated through simulation studies, and an application to a data set from the first HIV vaccine efficacy trial is provided.