Trustworthy Machine Learning
Causal and hypothesis-driven methods for robustness, including out-of-distribution detection.
I work on methods that make deep learning more reliable in biomedical imaging, computer vision, and NLP.
Current directions include:
- Out-of-distribution detection using hypothesis-driven and causal inference ideas
- A Rubin causal model–based stochastic proof-by-contradiction framework for model reliability
- Applications such as pathogen detection, lung-cancer cell profiling, and DNA damage screening