An Instrumental Variable Method for Point Processes with Application to Optogenetic Experiments
地点：Online (Tencent Meeting)
Abstract: Point processes are probabilistic tools for modeling event data such as neural spike trains, natural disasters, and crimes. While there exists a fast-growing literature studying the relationships between point processes, how such relationships connect to causal effects remains unexplored. In the presence of unmeasured confounders, parameters from point process models do not necessarily have causal interpretations. Motivated by the development of optogenetics in neuroscience, we propose an instrumental variable method for causal inference when both the treatment and the outcome are point processes.
In this talk, we examine causal quantities of interest and establish nonparametric identification results with a binary instrumental variable. We also propose an estimation strategy based on well-established deconvolution methods, which is applicable under many commonly-used models without requiring distributional assumptions on the unmeasured confounders. For empirical illustration, we apply the proposed methodology to the analysis of data from an optogenetic experiment that studies how neural circuits maintain stable odor representation in the mouse brain.
Bio: Shizhe Chen is an assistant professor in the Department of Statistics at University of California, Davis. He is an alumnus of Yuanpei College, and he obtained his Ph.D. in biostatistics at University of Washington. He was a postdoctoral research scientist at Department of Statistics and Grossman Center for the Statistics of Mind in Columbia University. His recent research focuses on development statistical theory and methods for neural data arising from modern neuroscience experiments.
ID: 199 197 803