Research
My work examines the societal implications of algorithmic decision-making.
Algorithms mediate many consequential decisions: who gets hired, who gets a loan, who sees which advertisement, what price a person is offered. I study how bias enters these processes, which mathematical notions of fairness are achievable, and what they cost in efficiency. The tools come mostly from combinatorial optimization, convex optimization, and online algorithms.
A recurring thread is the maximin objective — the Rawlsian principle that an outcome should be judged by how it treats the worst-off party. I've been looking at how it behaves in selection from posets, hypergraph partitioning, and facility location: problems interesting in their own right, and possibly a window into how such objectives behave across combinatorial optimization.
Peer-reviewed
Published
-
Algorithmic Challenges in Ensuring Fairness at the Time of Decision
-
Secretary Problems with Biased Evaluations using Partial Ordinal Information
-
Using Algorithms to Tame Discrimination: A Path to Algorithmic Diversity, Equity, and Inclusion
-
Don't Let Ricci v. DeStefano Hold You Back: A Bias-Aware Legal Solution to the Hiring Paradox
Working papers
In preparation
-
Expected Maximin Fairness in Max-Cut and Other Combinatorial Optimization Problems
-
Poset Selection in the Adversarial Setting
Chapters and expository writing
Other work
A full list of publications, talks, and service is in my curriculum vitae.