Thodoris Lykouris
I am an assistant professor of Operations Management at the MIT Sloan School of Management. I am broadly interested in machine learning, algorithms, optimization, and economics. In particular, my research focuses on data-driven sequential decision making.
Before joining MIT, I spent two years as a postdoctoral researcher at Microsoft Research NYC. I received my Ph.D. in the Computer Science department of Cornell University where I had the privilege to be advised by Éva Tardos. Prior to that, I received my Diploma in EECS from National Technical University of Athens. My Ph.D. research was supported in part by a Google Ph.D. Fellowship and a Cornell University Fellowship.
Contact
Email: lykouris at mit dot edu
Ph.D. Dissertation
- Effective online decision-making in complex multi-agent systems
Cornell University, August 2019
Finalist for the George B. Dantzig Dissertation Award (2020)
Journal articles
- Competitive caching with machine learned advice
- Small-loss bounds for online learning with partial information
- Pricing and optimization in shared vehicle systems: An approximation framework
with Sergei Vassilvitskii
Journal of the ACM (volume 68, issue 4, article 24, pp. 1-25, 2021): Journal version
Extended abstract appeared at ICML'18
Video from my talk at ICML
with Karthik Sridharan and Éva Tardos
Mathematics of Operations Research (forthcoming)
Extended abstract appeared at COLT'18
Video from my talk at COLT
Finalist in the INFORMS Nicholson Student Paper Competition (2018)
with Siddhartha Banerjee and Daniel Freund
Operations Research (forthcoming)
Extended abstract appeared at EC'17
Video from my talk at EC
Finalist in the Applied Probability Society Student Paper Competition (2017)
Working papers
- Contextual search in the presence of irrational agents
- Corruption-robust exploration in episodic reinforcement learning
- Static pricing for multi-unit prophet inequalities
- Stochastic bandits robust to adversarial corruptions
- Learning and efficiency in games with dynamic population
with Akshay Krishnamurthy, Chara Podimata, and Robert Schapire
Extended abstract to appear at STOC'21
Video from Chara's talk at STOC
with Max Simchowitz, Aleksandrs Slivkins, and Wen Sun
Extended abstract to appear at COLT'21
Video from my talk at Virtual RL theory seminar series
with Shuchi Chawla and Nikhil Devanur
with Vahab Mirrokni and Renato Paes Leme
Extended abstract appeared at STOC'18
Video from my talk at STOC
with Vasilis Syrgkanis and Éva Tardos
Extended abstract appearead at SODA'16
Conference publications
- Contextual search in the presence of irrational agents
- Corruption-robust exploration in episodic reinforcement learning
- Constrained episodic reinforcement learning in concave-convex and knapsack settings
- Bandits with adversarial scaling
- Feedback graph regret bounds for Thompson Sampling and UCB
- Advancing subgroup fairness via sleeping experts
- On preserving non-discrimination when combining expert advice
- Competitive caching with machine learned advice
- Stochastic bandits robust to adversarial corruptions
- Small-loss bounds for online learning with partial information
- Pricing and optimization in shared vehicle systems: An approximation framework
- Learning in games: Robustness of fast convergence
- Learning and efficiency in games with dynamic population
- Influence maximization in Switching-Selection Threshold Models
with Akshay Krishnamurthy, Chara Podimata, and Robert Schapire
53rd Annual ACM Symposium on the Theory of Computing (STOC 2021)
Journal version: Under preparation
with Max Simchowitz, Aleksandrs Slivkins, and Wen Sun
34th Annual Conference on Learning Theory (COLT 2021)
Journal version: Under preparation
with Kianté Brantley, Miroslav Dudik, Sobhan Miryoosefi, Max Simchowitz, Aleksandrs Slivkins, Wen Sun
34th Annual Conference on Neural Information Processing Systems (NeurIPS 2020)
with Vahab Mirrokni and Renato Paes Leme
37th International Conference on Machine Learning (ICML 2020)
Video from my talk at ICML
with Éva Tardos and Drishti Wali
31st International Conference on Algorithmic Learning Theory (ALT 2020)
with Avrim Blum
11th Innovations on Theoretical Computer Science Conference (ITCS 2020)
with Avrim Blum, Suriya Gunasekar, and Nathan Srebro
32nd Annual Conference on Neural Information Processing Systems (NeurIPS 2018)
with Sergei Vassilvitskii
35th International Conference on Machine Learning (ICML 2018)
Journal version: Journal of the ACM (forthcoming)
with Vahab Mirrokni and Renato Paes Leme
50th Annual ACM Symposium on the Theory of Computing (STOC 2018)
Journal version: Under preparation
with Karthik Sridharan and Éva Tardos
31st Annual Conference on Learning Theory (COLT 2018)
Journal version: Mathematics of Operations Research (forthcoming)
with Siddhartha Banerjee and Daniel Freund
18th ACM Conference on Economics and Computation (EC 2017)
Journal version: Operations Research (forthcoming)
with Dylan Foster, Zhiyuan Li, Karthik Sridharan, and Éva Tardos
30th Annual Conference on Neural Information Processing Systems (NeurIPS 2016)
Video from my talk at YoungEC
with Vasilis Syrgkanis and Éva Tardos
27th ACM-SIAM Symposium on Discrete Algorithms (SODA 2016)
Journal version: Under preparation
with Dimitris Fotakis, Evangelos Markakis, and Svetlana Obrazstova
7th International Symposium on Algorithmic Game Theory (SAGT 2014)