Nanlin Jin

Co-adaptive Strategies for Sequential Bargaining Problems with Discount Factors and Outside Options

Bargaining is fundamental in social activities. Game-theoretic methodology has provided theoretic solutions for certain abstract models. Even for a simple model, this method demands substantial human intelligent effort in order to solve game-theoretic equilibriums. The analytic complexity increases rapidly when more elements are included in the models. In our previous work, we have demonstrated how co-evolutionary algorithms can be used to find approximations to game-theoretic equilibriums of bargaining models that consider bargaining costs only. In this paper, we study more complicated bargaining models, in which outside option is taken into account besides bargaining cost. Empirical studies demonstrate that evolutionary algorithms are efficient in finding near-perfect solutions. Experimental results reflect the compound effects of discount factors and outside options upon bargaining outcomes. We argue that evolutionary algorithm is a practical tool for generating reasonably good strategies for complicated bargaining models beyond the capability of game theory.

A constraint-guided method with evolutionary algorithms for economic problems

This paper presents an evolutionary algorithms based constrain-guided method (CGM) that is capable of handling both hard and soft constraints in optimization problems. While searching for constraint-satisfied solutions, the method differentiates candidate solutions by assigning them with different fitness values, enabling favorite solutions to be distinguished more likely and more effectively from unfavored ones. We illustrate the use of CGM in solving two economic problems with optimization involved: (1) searching equilibriums for bargaining problems; (2) reducing the rate of failure in financial prediction problems. The efficacy of the proposed CGM is analyzed and compared with some other computational techniques, including a repair method and a penalty method for the problem (1), a linear classifier and three neural networks for the problem (2), respectively. Our studies here suggest that the evolutionary algorithms based CGM compares favorably against those computational approaches.

Bargaining strategies designed by evolutionary algorithms

This paper explores the possibility of using evolutionary algorithms (EAs) to automatically generate efficient and stable strategies for complicated bargaining problems. This idea is elaborated by means of case studies. We design artificial players whose learning and self-improving capabilities are powered by EAs, while neither game-theoretic knowledge nor human expertise in game theory is required. The experimental results show that a co-evolutionary algorithm (CO-EA) selects those solutions which are identical or statistically approximate to the known game-theoretic solutions. Moreover, these evolved solutions clearly demonstrate the key game-theoretic properties on efficiency and stability. The performance of CO-EA and that of a multi-objective evolutionary algorithm (MOEA) on the same problems are analyzed and compared. Our studies suggest that for real-world bargaining problems, EAs should automatically design bargaining strategies bearing the attractive properties of the solution concepts in game theory.

Incentive method to handle constraints in evolutionary algorithms with a case study

This paper introduces Incentive Method to handle both hard and soft constraints in an evolutionary algorithm for solving some multi-constraint optimization problems. The Incentive Method uses hard and soft constraints to help allocating heuristic search effort more effectively. The main idea is to modify the objective fitness function by awarding differential incentives according to the defined qualitative preferences, to solution sets which are divided by their satisfaction to constraints. It does not exclude the right to access search spaces that violate some or even all constraints. We test this technique through its application on generating solutions for a classic infinite-horizon extensive-form game. It is solved by an Evolutionary Algorithm incorporated by Incentive method. Experimental results are compared with results from a penalty method and from a non-constraint setting. Statistic analysis suggests that Incentive Method is more effective than the other two techniques for this specific problem.