In my most recent research, I developed a machine learning method for causal inference on observational data, fused extended two-way fixed effects (FETWFE). Uses for FETWFE include estimating the causal effect of a law or policy that is implemented state-by-state. My paper with my advisor Jacob Bien on estimating probabilities of rare events, like purchasing an item from a display ad, was accepted to the Fortieth International Conference on Machine Learning (ICML 2023). Finally, I also developed a method for stable feature selection with Prof. Bien. One application for our method is improving identification of DNA nucleotides associated with phenotypes in genome-wide association studies. These sorts of studies have been effective in applications ranging from providing targeted treatment for Hepatitis C to improving grain yields.
Publications
- Gregory Faletto and Jacob Bien (2023). Predicting Rare Events by Shrinking Towards Proportional Odds. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:9547-9602. [pdf] [preprint]
- Matthew J. Salganik, et al. Measuring the predictability of life outcomes with a scientific mass collaboration. Proceedings of the National Academy of Sciences 117.15 (2020): 8398-8403. [pdf]
Preprints
- Gregory Faletto (2023). Fused Extended Two-Way Fixed Effects for Difference-in-Differences with Staggered Adoptions. [pdf]
- Gregory Faletto and Jacob Bien (2022). Cluster Stability Selection. [pdf]
You can also see my research on my Google Scholar profile.