Aras Selvi
Imperial College London
Analytics & Operations, Imperial College Business School
Computational Optimization Group, Department of Computing
Contact: a[dot][last name]19@imperial.ac.uk
Address: ICBS E3.02, SW7 2AZ, London, UK
[Scholar] [LinkedIn] [GitHub] [X (Twitter)] [CV]

Welcome to my page!
I am a doctoral candidate at Imperial College Business School where I am fortunate to be advised by Wolfram Wiesemann as a member of the Analytics & Operations group. I am also affiliated with the Computational Optimization Group and the Data Science Institute of Imperial College London. I have recently completed a PhD placement at The Alan Turing Institute as a part of the Enrichment Scheme and a PhD AI Research Associate Internship at J.P. Morgan.
In Fall 2026, I will start as an Assistant Professor of Operations & Technology at the UCL School of Management. Prior to that, during the 2025-2026 academic year, I will be a Postdoctoral Research Associate at Princeton University ORFE, mentored by Bartolomeo Stellato!
My research interests are the theory of data-driven decision making under uncertainty and its applications in machine learning, privacy, and fairness. In my recent works, I have been working on designing optimal privacy mechanisms, developing efficient algorithms for robust machine learning, as well as approximating hard decision making problems via robust optimization.
News
Apr 2025
An extended abstract of our working paper "Optimal and fair housing allocation via weakly coupled Markov decision processes" is accepted for presentation at the MSOM Conference 2025 (London).Apr 2025
I will present a new, near-optimal mechanism for differential privacy at the Euro OSS on OR & ML.Apr 2025
We updated our work "It's all in the mix: Wasserstein classification and regression with mixed features". We substantially generalized our previous results, including the absence of a regularization reformulation of Wasserstein DRO when discrete features are present.