AI & Fundamentals

The Non-Stochastic Control Framework - Naman Agarwal, Research Scientist, Google AI, Princeton

Please register for this event here

 

Abstract:

Linear dynamical systems are a continuous subclass of reinforcement learning models that are widely used in robotics, finance, engineering, and meteorology. Classical control, has focused on dynamics with Gaussian i.i.d. Noise and quadratic loss functions in terms of provably efficient algorithms. I will present a non-stochastic control framework inspired by online learning, that generalizes some traditional notions of robust control and discuss methodology which achieves efficient control with adversarial noise and general convex loss functions.    

Bio:

Naman Agarwal is a Research Scientist at Google AI, Princeton. His research has primarily focussed on efficient algorithms for optimization in machine learning and online learning. Recently, he has been interested in bringing these ideas to control and reinforcement learning. He graduated with a PhD from Princeton University, under the supervision of Elad Hazan.

 

Please register for this event here