Aditya Bhaskara

About

I am an Associate Professor member at the Kahlert School of Computing at the University of Utah. I am also part of the Theory Group, and the Utah Center for Data Science. Prior to Utah, I spent a few years as a Postdoc at the Algorithms Group in Google Research NYC (2013-15), and at EPFL in Switzerland (2012-13). Going back further in time, I spent several years as a PhD student at Princeton University, and as an undergrad at IIT Bombay.

Selected publications

• Descent with Misaligned Gradients and Applications to Hidden Convexity (A. Bhaskara, A. Cutkosky, R. Kumar, M. Purohit) International Conference on Learning Representations (ICLR) 2025
• Convergence Guarantees for the DeepWalk Embedding on Block Models (C. Harker, A. Bhaskara) International Conference on Machine Learning (ICML) 2024
• New Tools for Smoothed Analysis: Least Singular Value Bounds for Random Matrices with Dependent Entries (A. Bhaskara, E. Evert, V. Srinivas, A. Vijayaraghavan) ACM Symposium on Theory of Computing (STOC) 2024
• Online Learning and Bandits with Queried Hints (A. Bhaskara, S. Gollapudi, S. Im, K. Kollias, K. Munagala) Innovations in Theoretical Computer Science (ITCS) 2023
• Additive Error Guarantees for Weighted Low Rank Approximation (A. Bhaskara, A.K. Ruwanpathirana, M. Wijewardena) International Conference on Machine Learning (ICML), 2021