Solving Combinatorial Games using Products, Projections and Lexicographically Optimal Bases
This work addresses computational bottlenecks in game theory for researchers and practitioners dealing with combinatorial games, offering incremental improvements in algorithm efficiency.
The paper tackles the problem of finding Nash equilibria in two-player zero-sum games with combinatorial strategies, such as spanning trees or matchings, by developing efficient algorithms for Bregman projections and multiplicative weights updates, resulting in polynomial-time simulations and single-projection solutions for symmetric equilibria.
In order to find Nash-equilibria for two-player zero-sum games where each player plays combinatorial objects like spanning trees, matchings etc, we consider two online learning algorithms: the online mirror descent (OMD) algorithm and the multiplicative weights update (MWU) algorithm. The OMD algorithm requires the computation of a certain Bregman projection, that has closed form solutions for simple convex sets like the Euclidean ball or the simplex. However, for general polyhedra one often needs to exploit the general machinery of convex optimization. We give a novel primal-style algorithm for computing Bregman projections on the base polytopes of polymatroids. Next, in the case of the MWU algorithm, although it scales logarithmically in the number of pure strategies or experts $N$ in terms of regret, the algorithm takes time polynomial in $N$; this especially becomes a problem when learning combinatorial objects. We give a general recipe to simulate the multiplicative weights update algorithm in time polynomial in their natural dimension. This is useful whenever there exists a polynomial time generalized counting oracle (even if approximate) over these objects. Finally, using the combinatorial structure of symmetric Nash-equilibria (SNE) when both players play bases of matroids, we show that these can be found with a single projection or convex minimization (without using online learning).