【新時代物流沙龍第58期】Adaptive Sampling-based Nonconvex and Nonsmooth approaches for Stochastic Programs with Implicitly Decision-dependent Uncertainty
時間:2024年11月14日(周四) 10:30-12:00 地點:思源東樓611 報告人簡介:劉俊驛,清華大學工業工程系準聘副教授。2019年于美國南加州大學獲得工業與系統工程博士學位。2023年入選國家級青年人才項目。目前研究方向為隨機優化,側重隨機優化與統計、機器學習的交叉研究。以第一作者身份在Operations Research, Mathematics of Operations Research, SIAM Journal on Optimization 等國際學術期刊上發表多篇文章。 報告摘要:We consider a class of stochastic programming problems where the implicitly decision-dependent random variable follows a nonparametric regression model with heteroskedastic error. To deal with the computational difficulty resulted from the latent decision dependency, we develop an adaptive sampling-based surrogate method that integrates the simulation scheme and statistical estimates in a way that the simulation process is adaptively guided by the algorithmic procedure.