Zhuo Yu's research interests are estimating functions in semiparametric models, censored data, causal inference in longitudinal studies.

Currently, he is collaborating with Mark van der Laan and Jennifer Bryan on a Marginal Structural Longitudinal Logistic Regression Model. This model can be used to measure the causal effect of time-dependent treatments or exposure with time-dependent covariates in epidemiology research. Following the theory of estimating functions in censored data models developed by Drs. van der Laan and Robins (Censored Data, 2000), he and his collaborators construct the class of estimating functions in such model. Their proposed One-Step estimators are far more efficient than the widely used IPAW estimator.

Zhuo also works on semiparametric regression models with Mark van der Laan, constructing the class of estimating functions in such models, in which the optimal estimating functions are obtained in closed form.