A list of my publications and presentations is available here.
My research interests:
My current primary research interest is in how scientists can learn and make use of causal knowledge, and I collaborate with clinicians and other domain scientists in order to use these methodological advances to help solve real-world problems. My specialty is machine learning methods that can either tolerate or explicitly model unmeasured common causes. I wrote my dissertation on the latter topic, available here. I primarily work with causal graphical models, but I’m also interested in statistical models more generally. In addition to my primary research topic, I have an active interest in, and do research on, the social aspects of scientific research.
Paper topics that I’m researching/writing about right now:
The conditions required for, and limitations of, automated methods for discovering hidden common causes
Demonstrating the methodological benefits of advanced causal graph learning methods over traditional methods for analyzing psychiatric data
Novel discoveries about the mechanisms of alcohol use disorder
The effects of, and inference of, unobserved variables for graphical models of time series data
The role of reproducibility in scientific research
A method for comparing graphs with spatially located variables
Evaluating the effectiveness of bootstrapping various causal graph learning methods