Quinn Winters is a rising junior at Case Western Reserve University in Cleveland, Ohio. He is interested in experimental results and their relation to the underlying deterministic theories. He plans to pursue graduate studies in mathematical probability or causal inference theory in the near future, and seeks to use simulations to describe and learn from non-scientific datasets (e.g., legal cases, economics, or rhetoric). In his free time, he likes reading philosophy and math books.
Project: Automated Causal Graph Learning for Statistical Inference
This project is centered around using causal graph generation methods to create better models and distributions to describe causal effects in a dataset. Using both doubly robust estimation (Funk et al, 2010), Markov-blanket learning (Pellet, 2008) and standard stabilized inverse probability weighting, the new Mathematica package will be able to learn a set of causal relationships from a dataset and produce a directed acyclic graph that models the interaction between variables in the dataset.
These algorithms will then be compared for efficiency, using RLink, and accuracy, using both artificial datasets and datasets used in previous epidemiological studies.