Abstract:We evaluate the best-response algorithm in the context of pure-integer convex quadratic games. We provide a sufficient condition that if certain interaction matrices (the product of the inverse of the positive definite matrix defining the convex quadratic terms and the matrix that connects one player's problem to another's) have all their singular values less than 1, then finite termination of the best-response algorithm is guaranteed regardless of the initial point. Termination is triggered through cycling among a finite number of strategies for each player. Our findings indicate that if cycling happens, a relaxed version of the Nash equilibrium can be calculated by identifying a Nash equilibrium of a smaller finite game. Conversely, we prove that if every singular value of the interaction matrices is greater than 1, the algorithm will diverge from a large family of initial points. In addition, we provide an infinite family of examples in which some of the singular values of the interaction matrices are greater than 1, cycling occurs, but any mixed-strategy with support in the strategies where cycling occurs has arbitrarily better deviations. Then, we perform computational tests of our algorithm and compare it with standard algorithms to solve such problems. We notice that our algorithm finds a Nash equilibrium correctly in every instance. Moreover, compared to a state-of-the art algorithm, our method shows similar performance in two-player games and significantly higher speed when involving three or more players.

On the Convexity of Independent Set Games

Convergence Dynamics of Graphical Congestion Games.

Independent Domination on Tree Convex Bipartite Graphs

Universally Balanced Combinatorial Optimization Games

Totally Balanced Combinatorial Optimization Games.

Algorithmic Aspects of the Core of Combinatorial Optimization Games

Optimization Games on Graphs(Continuous and Discrete Mathematical Optimization)

Impartial geodetic building games on graphs

A game theoretical formulation for distributed optimization problems

On Distributed Equilibrium Seeking for Generalized Convex Games

An efficient characterization of submodular spanning tree games

Independent Domination: Reductions from Circular- and Triad-Convex Bipartite Graphs to Convex Bipartite Graphs.

Linear and Integer Programming Techniques for Cooperative Games

On Equilibria in Constrained Generalized Games with the Weak Continuous Inclusion Property

Convex Discrete Optimization

Best-response Algorithms for Integer Convex Quadratic Simultaneous Games

Convex Markov Games: A Framework for Fairness, Imitation, and Creativity in Multi-Agent Learning

Algorithmic Cooperative Game Theory

Convex Combinatorial Optimization

Zero-sum Dynkin games under common and independent Poisson constraints

The Least-core and Nucleolus of Path Cooperative Games