Research Expertise and Interest

machine learning, personalized medicine and healthcare systems, global health equity, precision health

Research Description

Irene Chen is a machine learning researcher broadly interested in computational methods to improve healthcare for the entire patient population. She primarily works with electronic health records, as well as other patient data including disease registries, wearable data, and insurance claims. Her research has two main threads: 1) she develops machine learning methods for equitable clinical care, and 2) she audits, decomposes, and mitigates algorithmic bias. 

Current projects include health equity audits for large language models, disease progression modeling for diabetes, and adjusting for adverse effects of distribution shift on different patient subpopulations. Her collaborators include clinicians, health insurance providers, anthropologists, computer scientists, statisticians, and data scientists. Her work has been published widely in machine learning conferences (NeurIPS, AAAI), medical journals (Nature Medicine, Lancet Digital Health, JAMA), and venues for wider audiences (Quartz, Annual Reviews). 

Chen received her PhD from MIT EECS.  Before MIT, she received a joint AB/SM degree from Harvard University. She has also worked at Dropbox and Microsoft Research before joining UC Berkeley.