Jinyuan Liu

This paper introduces a semiparametric method for joint mean-variability inference in longitudinal studies with missing data, demonstrating robustness and reliability through simulations and a real-world application.

This paper introduces eCDF-copula, a method for generating synthetic longitudinal tabular data that preserves within- and between-visit dependence structures, outperforming existing methods in resemblance and utility.

The paper develops a regression framework to analyze longitudinal metabolomics data and mental wellness in breast cancer patients, identifying metabolites associated with mental health changes over time.

The paper develops a new estimator to more accurately assess machine learning model performance in predicting brain-phenotype associations and shows it reduces bias in estimating predictive accuracy.

This paper compares double machine learning and regularized partial correlation network methods for high-dimensional confounding in causal mediation analyses using a human microbiome study.

This study evaluates hybrid immunity against the Omicron variant JN.1 11 months after breakthrough infection, showing heterogeneous responses and increased neutralization following reinfection.