<ul><li>An optimization framework is developed focusing on transferring control authority to a parametric policy to create an autonomous dynamical system.</li><li>The framework allows optimizing policy parameters independently of controls or actions, without relying on approximate Dynamic Programming and Reinforcement Learning.</li><li>Simpler algorithms at the autonomous system level are derived, performing computations equivalent to policy gradients, Hessians, and other optimization methods.</li><li>The framework is applicable to various tasks like behavioral cloning, mechanism design, system identification, and tuning generative AI models.</li></ul>

Dynamical System Optimization

Discover more