Data Science Portfolio
Below is a selection of my open-source projects, statistical packages, and machine learning implementations. You can find all of my source code and contributions on my GitHub profile.
Applied Projects
Automated quality assurance of CT scans by training and validating a random forest model to identify
blood vessels with over 89% precision. Runs in a Docker container for end-to-end execution in under 1 minute.
Developed a frequentist hierarchical model to estimate the relative biological effectiveness (RBE) of proton vs. photon therapy in lung cancer, using a large-scale clinical dataset.
Statistical Packages
A Bayesian sum-of-trees model designed to incoporate measurement error in predictor
variables, with applications in medical and clinical research.
A Bayesian sum-of-trees model designed to handle semi-supervised learning problems with
multiple
instance learning (MIL) data structures, where observations are organized into bags and
only
bag-level labels are observed.
Methodological Projects
A model-agnostic ensemble method for making stronger predictions on clustered data for out-of-sample groups. Predictions are generated such that out-of-sample group predictions are more closely aligned with the most similar groups in the training data.
