I am a PhD student currently working with Prof Marco Carone and Prof Alex Luedtke. The topic is estimation of and inference for an overall summary of the distribution underlying the data under semiparametric/nonparametric models, especially when the summary involves a local feature (e.g. the density function or a regression function). Examples include average treatment effect, entropy of the distribution and variable importance. Typically we can construct root-n-consistent and asymptotically normal estimators for this summary of interest, based on which we can conduct statistical inference, but existing methods either require stringent smoothness assumptions or the knowledge of an influence function, which is hard to find in general, so their usage is usually limited to experts. We are trying to circumvent these difficulties for some problems with flexible sieve estimation.
I'm also doing research with Dr. Jennifer Bobb as an RA at Kaiser Permanente Washington Health Research Institute. The project we are working on is a pragmatic cluster-randomized trial on a clinic-level intervention treating opioid use disorder. Besides various exploratory analysis on preliminary (Phase I) data, I'm also involved in planning statistical analysis for the data from the Phase II trial.
I am a PhD student currently working with Prof Marco Carone and Prof Alex Luedtke. The topic is estimation of and inference for an overall summary of the distribution underlying the data under semiparametric/nonparametric models, especially when the summary involves a local feature (e.g. the density function or a regression function). Examples include average treatment effect, entropy of the distribution and variable importance. Typically we can construct root-n-consistent and asymptotically normal estimators for this summary of interest, based on which we can conduct statistical inference, but existing methods either require stringent smoothness assumptions or the knowledge of an influence function, which is hard to find in general, so their usage is usually limited to experts. We are trying to circumvent these difficulties for some problems with flexible sieve estimation.
I'm also doing research with Dr. Jennifer Bobb as an RA at Kaiser Permanente Washington Health Research Institute. The project we are working on is a pragmatic cluster-randomized trial on a clinic-level intervention treating opioid use disorder. Besides various exploratory analysis on preliminary (Phase I) data, I'm also involved in planning statistical analysis for the data from the Phase II trial.