My research has touched on several areas, including Kidney Exchange, Blood Donation, Preference Elicitation for Policy
Design, and how to design and apply these tools in a responsible way.
Patients with kidney failure have only two options: a lifetime on dialysis, or kidney
transplantation. Dialysis is far more expensive and burdensome than transplantation, however
donor kidneys are scarce — on average, 20 people die each day in the US while waiting for a transplant.
Furthermore, many patients in need of a kidney have willing living donors, but cannot undergo
transplantation due to medical incompatibilities.
To address this supply-demand mismatch, kidney exchange allows patients with willing living
donors to swap their donors in order to find a compatible (or better) patient-donor match.
Formulated as an optimization problem, kidney exchange is NP-hard and APX-hard, though modern
exchanges are solvable in a reasonable amount of time (due to efficient formulations such as PICEF and PC-TSP).
In addition to being computationally hard, kidney exchange raises several logistical and ethical
challenges. My research has touched on several of these challenges, including: fairness for
marginalized patients, robustness to uncertainty, and moral considerations of designing a kidney
Robustness to Uncertainty
There are many sources of uncertainty in real kidney exchanges — due to medical, moral, and policy
factors. These sources of uncertainty are difficult to characterize, and can severely impact the
outcome of an exchange. We investigate techniques from robust optimization to address this problem.
- McElfresh, Duncan C, Michael Curry, Tuomas Sandholm,
and John P Dickerson, "Improving Policy-Constrained Kidney Exchange via Pre-Screening.”
Advances in Neural Information Processing Systems 33: Annual Conference on Neural
Information Processing Systems (NeurIPS), 2020.
- Bidkhori, Hoda, John P Dickerson, Ke Ren, and Duncan C
McElfresh. “Kidney exchange with Inhomogeneous Edge Existence Uncertainty.”
Conference on Uncertainty in Artificial Intelligence (UAI). 2020.
- McElfresh, Duncan C, Hoda Bidkhori, and John P
Dickerson, "Scalable Robust Kidney Exchange.” Conference on Artificial Intelligence
(AAAI), 2019. (pdf)
Fairness in Kidney Exchange
How can we prioritize marginalized patients, without severely impacting the overall exchange? We
study several different methods for enforcing this notion of fairness, and demonstrate their
effects on data collected from real exchanges.
- McElfresh, Duncan C, and John P Dickerson,
"Balancing lexicographic fairness and a utilitarian objective with application to kidney
exchange." Conference on Artificial Intelligence (AAAI), 2018. (pdf)
Ethics and Kidney Exchange
Designing a kidney exchange program requires input from medical professionals, policymakers,
computer scientists, and ethicists. A "good" program should be both technically- and
morally-sound — however technical experts (e.g., computer scientists) and stakeholders (e.g.,
medical professionals) often work independently. We propose a formal division of labor between
technical experts and stakeholders, and outline a framework through which these experts can
collaborate. Through this framework we analyze existing kidney exchange programs and survey the
technical literature on kidney exchange algorithms. We identify areas for future collaboration
between technical experts and stakeholders.
McElfresh, Duncan C, Vincent Conitzer, and John P Dickerson. “Ethics and Mechanism
Design in Kidney Exchange.” (Working paper).
Presentation: McElfresh, Duncan C, Patricia Mayer, Gabriel Schnickel, and
John P Dickerson. "Ok Google: Who Gets the Kidney?": Artificial Intelligence and
Transplant Algorithms. Panel presentation and discussion at the annual meeting of
the American Society of Bioethics and Humanities (ASBH), Anaheim, CA. (slides)
Blood is a scarce resource that can save the lives of those in need, and managing the blood
supply chain has been a topic of research for decades. We consider an aspect of the blood supply
chain that is seldom addressed by the literature: coordinating a network of donors to meet
demand from a network of recipients. The advent of massive online networks presents an
unprecedented opportunity to increase the number and impact of blood donations.
McElfresh, Duncan C, Christian Kroer, Sergey Pupyrev, Eric
Sodomka, Karthik Abinav Sankararaman, Zack Chauvin, Neil Dexter, John P Dickerson. “Matching
Algorithms for Blood Donation” The 21st ACM Conference on Economics and Computation
(EC). 2020. (link)
Poster and Presentation: McElfresh, Duncan C, Christian Kroer, Sergey
Pupyrev, Eric Sodomka, John P Dickerson. “Matching Algorithms for Blood Donation.”
Workshop on Mechanism Design for Social Good MD4SG, 2019.
for Policy Design
Preference elicitation is concerned with figuring out what people want, by asking carefully
selected questions. This field has a rich literature with roots in marketing, auctions, and
finance — among other applications. We use preference elicitation for a different purpose: to
design algorithms and policies that respect stakeholder interests. We are currently developing
elicitation schemes to identify policy priorities in kidney exchange and public
Phebe Vayanos, Duncan C McElfresh, Yingxiao Ye, John P Dickerson, and Eric Rice.
“Active Preference Elicitation via Adjustable Robust Optimization.” (pdf) (Under review at Management Science.)
Presentation: McElfresh, Duncan C, Phebe Vayanos, Eric Rice, and John P
Dickerson. “Optimizing Public Policy for Homelessness Assistance.” INFORMS Annual
Presentation: McElfresh, Duncan C, Phebe Vayanos, John P Dickerson. “Robust
Active Preference Elicitation, for Learning Policy Priorities.” Presentation at 2019 INFORMS Revenue
Management & Pricing Workshop.
AI applications in resource allocation and market design can have a measurable positive impact
on society; indeed, in the case of kidney exchange and blood donations, AI-powered tools can
literally help save lives. These AI applications can also have unintended consequences. Some of
my work has focused on understanding these consequences. For example: How does an AI-generated
suggestion impact a decision-maker's behavior? Does the general public agree with, or
understand, the notions of fairness or bias defined by computer scientists?
McElfresh, Duncan C, Lok Chan, Kenzie Doyle, Walter Sinnott-Armstrong, Vincent
Conitzer, Jana Schaich Borg,John P Dickerson. “Indecision Modeling.”
Conference on Artificial Intelligence (AAAI). 2021. (pdf) (poster) [arXiv]
Saha, Debjani, Candice Schumann, Duncan C McElfresh, John P
Dickerson, Michelle L Mazurek and Michael Carl Tschantz. “Measuring Non-Expert Comprehension
of Machine Learning Fairness Metrics.” Proceedings of the Thirty-seventh International
Conference on Machine Learning (ICML). 2020.
Chan, Lok, Kenzie Doyle, Duncan C McElfresh, Vincent Conitzer, John P Dickerson, Jana
Schaich Borg and Walter Sinnott-Armstrong. “Artificial Artificial Intelligence: Measuring
Influence of AI "Assessments" on Moral Decision-Making.” AAAI/ACM Conference on
Artificial Intelligence, Ethics, and Society (AIES),
Saha, Debjani, Candice Schumann, Duncan C McElfresh,, John P Dickerson, Michelle L
Michael Carl Tschantz. “Human Comprehension of Fairness in Machine Learning." AAAI/ACM
Conference on Artificial Intelligence, Ethics, and Society (AIES),
- McElfresh, Duncan C, Vincent Conitzer, and John P
Dickerson. “Ethics and Mechanism Design in Kidney Exchange.” (Working paper.)
McElfresh, Duncan C, Samuel Dooley, Charles Cui, Kendra Griesman, Weiqin Wang, Tyler
Will, Neil Sehgal and John Dickerson. “Can an Algorithm be My Healthcare Proxy?” 2020
International Workshop on Health Intelligence (at AAAI). 2020. (Workshop Paper.)
McElfresh, Duncan C. “A Framework for Technically- and Morally-Sound AI.”
AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AIES), 2019.
(Student program and poster.)