I am the Harold R. and Mary Anne Nash Early Career Professor and Assistant Professor in the School of Electrical and Computer Engineering and H. Milton Stewart School of Industrial and Systems Engineering at Georgia Institute of Technology. I received the B.Tech (with honors) degree from the Indian Institute of Technology, Madras and the Ph.D. degree in Electrical Engineering from University of California, Berkeley. Before joining Georgia Tech, I spent a semester at the Simons Institute for the Theory of Computing as a research fellow for the program “Theory of Reinforcement Learning.”
My broad interests are in game theory, online and statistical learning. I am particularly interested in designing learning algorithms that provably adapt in strategic environments, fundamental properties of overparameterized models, and the foundations of multi-agent decision-making. In my spare time, I enjoy singing Carnatic vocal music, playing the piano, and long-distance cycling.
See here for a more formal bio in the third person.
Recent News
-
- November 2024: Our work on optimal estimation of missing mass in a stationary Markov sequence (joint with Ashwin Pananjady and Andrew Thangaraj) was published in Journal of Machine Learning Research.
- November 2024: Congratulations to Kuo-Wei Lai for receiving the ARC-ACO Fellowship (Spring 2025 cycle)!
- November 2024: Our work on the impossibility of convergence of mixed strategies arising from no-regret learning (joint with Soham Rajesh-Phade and Anant Sahai) was published in Mathematics of Operations Research.
- October 2024: I presented our work on comparison and transfer between tasks in overparameterized learning at the Institute for Pure and Applied Mathematics Workshop on “Theory and Practice of Deep Learning”.
- August 2024: Our work on equivalences between the kernel SVM and minimum-Hilbert-norm interpolation, led by Chiraag Kaushik and Andrew McRae, was published in SIAM Journal on Mathematics of Data Science. Congratulations, Chiraag and Andrew!
- July 2024: My student Chiraag Kaushik presented our work on spectral imbalance correlating with class-wise accuracy disparities on pretrained models at ICML 2024. Congratulations, Chiraag!
- July 2024: My student Kuo-Wei Lai is presenting our work on sharp out-of-distribution error analysis of importance-weighted classifiers at IEEE International Symposium on Information Theory 2024. Congratulations, Kuo-Wei!
- June 2024: I presented our work on the algorithmic complexity of infinite-horizon general-sum stochastic games at the Dagstuhl workshop on Stochastic Games.
- April 2024: Our work on the good, bad and ugly sides of data augmentation, led by Chi-Heng (Henry) Lin and Chiraag Kaushik, is published at Journal of Machine Learning Research. Congratulations, Henry and Chiraag!
- Feb 2024: I’m chairing the second ITALT symposium that will be co-located with the annual Information Theory and Applications Workshop, and the Algorithmic Learning Theory Conference in San Diego! Details about the symposium here.
- Dec 2023: My students Guanghui Wang and Tyler LaBonte presented their papers on faster margin maximization rates for generic optimization methods, and towards last-layer retraining for group robustness with fewer annotations at NeurIPS 2023. Congratulations, Guanghui and Tyler!
- Dec 2023: I will be co-presenting a tutorial at NeurIPS 2023 on Dec 11 on “Reconsidering overfitting in the age of overparameterized models” with Spencer Frei and Fanny Yang. Schedule of all tutorials here!