时间：2019年05月09日（星期四）16：00-17：00

地点：沙河校区，主教楼 402

报告题目： A Global Bias-Correction DC Method forBiased Estimation under Memory Constraint

报告人：林路，山东大学金融研究院教授、博士生导师、副院长

报告摘要：This paper introduces aglobal bias-correction divide-and-conquer (GBC-DC) method for biased estimationunder the case of memory constraint. In order to introduce the new estimation,a closed representation of the local estimators obtained by the data in each batchis adopted to formulate a pro forma linear regression between the localestimators and the true parameter of interest. A least squares is used withinthis framework to composite a global estimator of the parameter. Thus, the maindifference from the classical DC method is that the new GBC-DC method can absorbthe information hidden in the statistical structure and the variables in eachbatch of data. Consequently, the resulting global estimator is strictlyunbiased even if the local estimators have a non-negligible bias. Moreover, theglobal estimator is consistent under some mild conditions, and even can achieveroot-$n$ consistency when the number of batches is large. The new method issimple and computationally efficient, without use of any iterative algorithmand local bias-correction. Moreover, the proposed GBC-DC method applies tovarious biased estimations such as shrinkage-type estimation and nonparametricregression estimation. Based on our comprehensive simulation studies, theproposed GBC-DC approach is significantly bias-corrected, and the behavior iscomparable with that of the full data estimation.

报告人简介：林路是山东大学金融研究院教授、博士生导师、副院长；在南开大学获得博士学位后，先在南开大学任教，然后到山东大学任教至今；从事高维统计、非参数和半参数统计以及金融统计等方的研究，在国际统计学、机器学习和相关应用学科顶级期刊Annals of Statistics, Journal of MachineLearning Research, PLoS computational biology和其它重要期刊发表研究论文100余篇；主持过多项国家自然科学基金课题、博士点专项基金课题、山东省自然科学基金重点项目等；获得国家统计局颁发的统计科技进步一二等奖，山东省优秀教学成果一等奖；是国家973项目、国家创新群体和教育部创新团队的核心成员，教育部应用统计专业硕士教育指导委员会成员，山东省政府参事。

本次活动受中央财经大学2019专题学术讲座项目资助。