Second Order Methods for Bandit Optimization and Control

Explore advanced optimization techniques in this 56-minute conference talk that delves into second-order methods for bandit optimization and control problems. Learn how these sophisticated mathematical approaches can improve decision-making in uncertain environments where feedback is limited or delayed. Discover the theoretical foundations and practical applications of second-order optimization methods, including their advantages over first-order approaches in bandit settings. Understand how these methods can be applied to control systems and sequential decision-making problems. Examine the convergence properties, computational considerations, and performance guarantees of second-order bandit algorithms. Gain insights into cutting-edge research that bridges optimization theory with reinforcement learning and control theory, presented as part of the Data Science: Probabilistic and Optimization Methods program at the International Centre for Theoretical Sciences.