Recent Thesis Topics

2017

  • An algorithm for identifying naturalistic boundaries in robot-derived kinematic data and the efficacy of sub-setting in modeling Fugl-Meyer Assessments
  • Recidivism of Women Living with HIV and at-risk for HIV on Probation: A Recurrent Events Survival Analysis
  • Critical Window Variable Selection Modeling for Air Pollution and Birth Outcome
  • Identifying Disease Heterogeneity from Gene Expression Data by Integrating Prior Pathway Information using Dirichlet Process
  • Assessing non-coding genome annotations through enrichment analysis of 15 genome-wide association studies
  • Longitudinal Associations of Gay and Bisexual Men's Stress and Mental Health
  • Competing Risk Models for Assessing Mortality Among Men with Prostate Cancer

2016

 

  • Analysis of ART Drug Change on Advanced AIDS Patients Clinical Outcomes
  • Analysis of Hospice Use in Determining Medical Expenditures in the SEERMedicare Data
  • A Method for Improving Goodness of Fit in Multi-Directional Kinematic Data Collected via Planar Robots
  • Unsupervised Pathway Based Clustering Analysis on Gene Expression Data
  • Improving Disease Risk Prediction Through Integrating Tunctional Annotation 
  • Quantification of Relative Selection Importance of Gene Mutations in Cholangiocarcinoma Subtypes using Cancer Selection Intensity Model-Averaged Clustering (CSI-MAC)
  • Spatial-Temporal Models for Human Lyme Disease in Connecticut
  • Exposure, Hazard, and the Temporal Dynamics of Diffusion on Social Networks
  • Longitudinal Analysis of Serial Measurements of Prostate-Specific Antigen (PSA)
  • Semi-supervised Learning: a New Method and Applications in Biomedical Studies
  • Penalized Regression Models for Genetic Risk Prediction using GWAS Data
  • Gender Difference in the Efficacy of Smoking Cessation Therapies 

2015

  • Classification of risk factors for mortality and morbidity after pulmonary resection
  • Evaluating the Effectiveness of Individual Placement and Support (IPS) Model of Supported Employment in Observational Studies
  • Incorporating ENCODE data in the analysis of genome wide association data to improve replication rates
  • Modeling nonignorable dropouts in longitudinal studies to alleviate attrition bias: An illusion of instrumentl variable approach