This is home for my research publications.
EHR Data Analysis: I have multiple years of experience in designing and implementing statistical models for research studies that utilize electronic health record (EHR) data to determine risk factors for clinical outcomes, identifying high risk subgroups, and predicting clinical outcomes.
Topic Modeling: My research on topic modeling explored latent topics underlying text data and developing new approaches for extracting information from examinees’ written responses.
Longitudinal Data Analysis: I have conducted several studies on latent variable modeling and longitudinal data analysis.
This study introduces a supervised topic model that incorporates finite-mixture modelling into supervised latent Dirichlet allocation. This model can detect latent groups of participants that have different relationships between their textual responses and associated scores.
This study considers Bayesian cross-domain latent growth curve models within a structural equation modeling framework to model dynamically changing relationships between proportion body fat and body mass index.
This study used interrupted time series analysis to compare mental health visits and psychopharmaceutical prescribing before and after the COVID-19 onset.
This study evaluated the relationship between sociodemographic factor including family structure and mental health service utilization before and during the COVID-19 pandemic.
The goal of this study is to define a set of PASC conditions, and to describe the timing of the conditions, by applying to a diverse population and comparison group of similar PCR-negative individuals.