报告题目：Resampling Strategy in Sequential Monte Carlo for Constrained Sampling Problems

主 讲 人：林明

报告时间：12月1日（周三）9:00-10:00

报告地点：腾讯会议 ID：647 149 161

报告摘要：

Monte Carlo sample paths of a dynamic system are useful for studying the underlying system and making statistical inferences related to the system. In many applications the dynamic system under study involves various types of constraints or observable features that need to be incorporated. In this paper we investigate effcient methods for generating sample paths of dynamic systems with rare and strong constraints, under a sequential Monte Carlo (SMC) framework. Specifically, we present a general formulation of the constrained sampling problem. Under the formulation, we propose a flexible resampling strategy based on a potentially time-varying lookahead timescale and identify the corresponding optimal resampling priority scores, for solving the constrained sampling problem. Two efficient approaches for estimating the optimal priority scores are developed, using forward pilots and backward pilots correspondingly. Several examples are used to illustrate the performance of the proposed methods.

主讲人简介：

林明，厦门大学王亚南经济研究院、经济学院统计学与数据科学系教授，全国应用统计专业学位研究生教育指导委员会委员，福建省统计科学重点实验室副主任。研究方向为政策评估方法和理论、贝叶斯统计、蒙特卡罗方法，研究成果发表在Journal of the American Statistical Association、Journal of Business & Economic Statistics、Statistical Science、《统计研究》等期刊上，现主持1项国家自然科学基金重点项目。

邀请人：林路

欢迎各位老师同学积极参加！