Independence weights for causal inference with continuous treatments
作者:
时间:2025-09-22
阅读量:152次
  • 演讲人: 陈冠华(University of Wisconsin–Madison, Associate Professor)
  • 时间:2025年9月26日15:00
  • 地点:浙江大学紫金港校区行政楼1417报告厅
  • 主办单位:浙江大学数据科学研究中心

Abstract: Studying causal effects of continuous treatments is important for gaining a deeper understanding of many interventions, policies, or medications, yet researchers are often left with observational studies for doing so. In the observational setting, confounding is a barrier to the estimation of causal effects. Weighting approaches seek to control for confounding by reweighting samples so that confounders are comparable across different treatment values. Yet, for continuous treatments, weighting methods are highly sensitive to model misspecification. In this talk, we elucidate the key property that makes weights effective in estimating causal quantities involving continuous treatments. We show that to eliminate confounding, weights should make treatment and confounders independent on the weighted scale. We develop a measure that characterizes the degree to which a set of weights induces such independence. Further, we propose a new model-free method for weight estimation by optimizing our measure. We study the theoretical properties of our measure and our weights, and prove that our weights can explicitly mitigate treatment-confounder dependence. The empirical effectiveness of our approach is demonstrated in a suite of challenging numerical experiments, where we find that our weights are quite robust and work well under a broad range of settings. Lastly, in an ongoing work, we also utilize new theoretical approaches for showing uniform convergence of our optimization-based weights to the true weights.


Bio: Guanhua Chen, Ph.D. is an Associate Professor in the Department of Biostatistics and Medical Informatics at the University of Wisconsin–Madison. He earned his Ph.D. in Biostatistics from the University of North Carolina at Chapel Hill and his B.S. in Bioinformatics from Huazhong University of Science and Technology. Dr. Chen’s research focuses on developing advanced statistical learning and causal inference methods to advance precision medicine, with applications spanning high-dimensional biomedical data such as genomics, microbiome, electronic health records, and other complex health datasets. Dr. Chen has served as principal investigator on multiple projects funded by the National Institutes of Health (NIH), the National Science Foundation (NSF), and the Patient-Centered Outcomes Research Institute (PCORI), with total funding exceeding five million USD. He has published nearly one hundred papers in leading journals, including Journal of the American Statistical Association, Biometrika, Biostatistics, Biometrics, Bioinformatics, Genome Biology, and Journal of American Medical Informatics Association.

陈冠华,博士,2014年博士毕业于美国北卡罗来纳大学教堂山分校,目前任美国威斯康星大学麦迪逊分校(University of Wisconsin–Madison)生物统计与医学信息学系副教授。陈博士的研究方向集中于统计学习和因果推断方法学的开发与应用,以推动精准医学的发展。他的工作涵盖高维和复杂生物医学数据,包括基因组学、微生物组、电子健康记录 及其他临床大数据。在科研成果方面,陈教授主持和参与的研究获得了美国国立卫生研究院(NIH)、国家科学基金会(NSF)、病人中心结局研究院(PCORI)等多项资助,总经费超过五百万美元。他已在Journal of the American Statistical Association、Biometrika、Biostatistics、Genome Biology、JAMA Network Open等国际权威期刊发表论文近百篇,研究成果在因果推断与精准医学领域具有一定影响。