Bio and CV
Elad Hazan is a professor of computer science at Princeton University and director of Princeton AI Alignment and Safety (AIAS). He studies the foundations of machine learning, optimization, and control, with a current focus on evaluating and aligning increasingly capable AI using online learning, game theory, and mechanism design.
He co-developed AdaGrad, an adaptive optimization algorithm widely used in machine learning; developed sublinear-time convex optimization algorithms; and helped establish online nonstochastic control theory. He also co-founded Google AI Princeton.
An ACM Fellow, Hazan has received the Bell Labs Prize and two IBM Goldberg Best Paper Awards. His research has been supported by a European Research Council grant, a Marie Curie fellowship, and a Google Research Award.
Research highlights
- Online Newton Step: logarithmic regret for exp-concave losses
- First sublinear-time algorithms for machine learning and semidefinite programming
- AdaGrad: adaptive gradient methods
- Fast, simple low-rank SDP solver with guarantees
- Linearly convergent conditional gradients for smooth, strongly convex optimization over polytopes
- Introducing nonstochastic control theory
More details and papers: group webpage.