Structure of Constraints in Sequential Decision-Making

The workshop will explore the impact of problem geometry on algorithmic performance in decision-making processes. Example topics include resource-constrained decision-making settings, the interaction between state-space geometry and algorithmic guarantees (e.g., downward-closed and matroid settings, discrete vs. continuous, convexity, network structure), the structure of transitions (stationary vs. non-stationary, exogenous randomness, correlation), and more complex induced structures (Lipschitz/convexity of value functions, induced problem dimensions). In addition, the workshop will also explore constraints arising from more practical concerns (simplicity, distributed control), as well as from incentives in dynamic mechanisms.