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. Time permitting, I will also discuss applications of these ideas to planning.