Tuesday, February 7, 2017
3:30 pm - 4:30 pm
701 Blockley Hall
Title: "Learning from Large Healthcare Data"
Abstract: With the advancement of modern technologies, biomedical research nowadays is facilitated with a variety of new types of data, e.g., mobile health (mHealth) data, electronic health records (EHR), and data from safety surveillance monitoring systems. Each of them has unique features and brings unique challenges to modern statisticians. For example, the mHealth data contain tremendous information that provides unprecedented opportunities for novel intervention of complex behaviors, e.g., smoking cessation, and have great promise of dynamic intervention/prevention and health management that is integrated with daily life. To design effective intervention program, it is crucial to develop statistical methods that are cable of capturing the dynamic process of human behaviors and change of health status over time.
The primary goal of my research is to develop statistical and computational methods that are intuitive, efficient and robust in real applications. In this talk, I would like to share my own experience on how and what we can learn from large healthcare data. I will talk about a Bayesian dynamic mediation analysis framework that enables researchers to capture the dynamic nature of the mediation process and to identify population heterogeneity.