Stephen Pfohl

I am a research scientist at Google and a recent graduate of the Biomedical Informatics PhD program at Stanford University. My work focuses on understanding fairness, robustness, and transparent evaluation of systems that use machine learning to inform clinical decision making.

Select Publications

Net benefit, calibration, threshold selection, and training objectives for algorithmic fairness in healthcare
Stephen R. Pfohl, Yizhe Xu, Agata Foryciarz, Nikolaos Ignatiadis, Julian Genkins, Nigam H. Shah
ACM Conference on Fairness Accountability and Transparency (FAccT) 2022
[preprint] [paper] [pdf]

Recommendations for algorithmic fairness assessments of predictive models in healthcare: evidence from large-scale empirical analyses
Stephen R. Pfohl
[dissertation] [pdf] [slides] CC BY 4.0

Evaluating algorithmic fairness in the presence of clinical guidelines: the case of atherosclerotic cardiovascular disease risk estimation
Agata Foryciarz, Stephen R. Pfohl, Birju Patel, Nigam H. Shah
BMJ Health & Care Informatics 29, e100460
[preprint] [paper]

A comparison of approaches to improve worst-case predictive model performance over patient subpopulations
Stephen R. Pfohl, Haoran Zhang, Yizhe Xu, Agata Foryciarz, Marzyeh Ghassemi, Nigam H. Shah.
Scientific Reports 12 (1), 1-13, 2022
[paper] [preprint] [code]

An Empirical Characterization of Fair Machine Learning For Clinical Risk Prediction
Stephen R. Pfohl, Agata Foryciarz, Nigam H. Shah.
Journal of Biomedical Informatics, 113:103621, 2021
[paper] [preprint] [code]

Counterfactual Reasoning for Fair Clinical Risk Prediction
Stephen R. Pfohl, Tony Duan, Daisy Yi Ding, Nigam H. Shah.
Proceedings of the 4th Machine Learning for Healthcare Conference, PMLR 106:325-358, 2019
[abstract] [pdf]

Creating fair models of atherosclerotic cardiovascular disease risk
Stephen Pfohl, Ben Marafino, Adrien Coulet, Fatima Rodriguez, Latha Palaniappan, Nigam H Shah.
Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society, 2019
[arxiv post-print with erratum] [pdf] [paper]