Online Decision Making

Online decision-making research focuses on developing algorithms that make optimal choices sequentially, adapting to changing environments and limited information. Current work emphasizes robust algorithms like Thompson Sampling and contextual bandits, often incorporating graph neural networks for complex data structures or adapting gradient descent for cost-sensitive scenarios. These advancements are crucial for applications ranging from personalized digital interventions and autonomous robotics to resource allocation in dynamic systems like energy trading, improving efficiency and safety in diverse fields. The field's impact stems from its ability to create adaptive and efficient systems that learn from experience in real-time.