Particle Swarm Optimization Approaches to Coevolve Strategies for the Iterated Prisoner's Dilemma

Franken, N.; Engelbrecht, Andries P.

doi:10.1109/tevc.2005.856202

Cited by 103 publications

(37 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Nelis F et al en [30], [31] presentan la aplicación de técnicas de coevolución basados en Particle Swarm Optimization (PSO) para evolucionar estrategias en el dilema del prisionero iterado y se usan tres diferentes técnicas de coevolución: una red neuronal entrenada con PSO, que se utiliza para predecir la próxima acción para ejecutar, un enfoque binario del PSO en el que la partícula representa un estrategia de juego completa y, finalmente, un enfoque que explora las estructuras simétricas de las estrategias creadas por los humanos. Los resultados experimentales indican que el enfoque de redes neuronales es capaz de inducir altos niveles de cooperación; sin embargo, en algunas ocasiones, podrían conducir a colapsos catastróficos por el despliegue de estrategias débiles.…”

Section: Evolución De Reglas Y Comportamien-tosunclassified

Economía Computacional Basada en Agentes

Giraldo-Giraldo

Gómez-Perdomo²

2012

Lámpsakos

View full text Add to dashboard Cite

Resumen. El artículo tiene como objetivo mostrar varios trabajos de investigación sobre un enfoque de simulación denominado Economía computacional basada en agentes, el cual rechaza las asunciones de los enfoques de estudio tradicionales que indican que la economía es un sistema cerrado que eventualmente logra un estado de equilibrio, en el que deben realizarse supuestos de racionalidad perfecta e inversiones homogéneas para que los modelos sean tratados analíticamente. En su lugar, ve a la economía como un sistema complejo, adaptativo y dinámico. Este nuevo enfoque permite usar la simulación basada en agentes para comprender que varios agentes econó-micos (firmas, grupos económicos) con sus propias reglas y objetivos, son capaces de interactuar entre sí y con su entorno para obtener comportamientos emergentes que no son explicables directamente de las propiedades de los agentes individuales.Palabras clave: Economía; Evolución de Reglas; Modelamiento; Simulación.Abstract. This article aims to show different research that have been developed on simulation approach called, agent-based computational economics, which rejects the assumptions of traditional approaches that indicate that the economy is a closed system eventually achieves a state equilibrium, in which assumptions of rationality perfect, homogeneous investments, among others, be made to allow the models are analytically tractable, instead, sees the economy as a complex adaptive system and dynamic.This new approach allows the use of agent-based simulation for the purpose of understand how various economic agents (firms, business groups) with their own rules and objectives are able to interact with each other and with their environment, resulting emergent behaviors not end be explained directly from the properties of the individual agents.

show abstract

Section: Evolución De Reglas Y Comportamien-tosunclassified

Economía Computacional Basada en Agentes

Giraldo-Giraldo

Gómez-Perdomo²

2012

Lámpsakos

View full text Add to dashboard Cite

show abstract

“…The algorithm is based on the flocking behavior of fish/birds, and has previously been successfully applied to many optimization as well as some gaming problems [1,2,4,7]. Similar to HC, each particle uses a set of updating equations to direct itself and the rest of the swarm toward the optimum.…”

Section: Particle Swarm Optimizationmentioning

confidence: 99%

Potential AI Strategies to Solve the Commons Game: A Position Paper

Verkhogliad

Oommen

2010

Advances in Artificial Intelligence

View full text Add to dashboard Cite

Abstract. In this paper, we propose the use of hill climbing and particle swarm optimization to find strategies in order to play the Commons Game (CG). The game, which is a non-trivial N -person non-zero-sum game, presents a simple mechanism to formulate how different parties can use shared resources. If the parties cooperate, the resources are sustainable. However, the resources get depleted if used indiscriminately. We consider the case when a single player has to determine the "optimal" solution, and when the other N − 1 players play the game by choosing the options with a fixed probability vector.

show abstract

“…For the IPD games involving deterministic and reactive strategies, several representations have been investigated in the past, which include a lookup table with bit-string encoding [13], finite-state machines [55]- [57], and neural networks [50], [53], [62], [63]. In this paper, we use the direct lookup table strategy representation that we have introduced and studied in [50], which directly represents IPD strategy behaviors through a one-to-one mapping between the genotype space (strategy representation) and the phenotype space (behaviors).…”

Section: B Strategy Representationmentioning

confidence: 99%

Relationship Between Generalization and Diversity in Coevolutionary Learning

Chong

Tiňo

Yao

2009

IEEE Trans. Comput. Intell. AI Games

View full text Add to dashboard Cite

Abstract-Games have long played an important role in the development and understanding of coevolutionary learning systems. In particular, the search process in coevolutionary learning is guided by strategic interactions between solutions in the population, which can be naturally framed as game playing. We study two important issues in coevolutionary learning-generalization performance and diversity-using games. The first one is concerned with the coevolutionary learning of strategies with high generalization performance, that is, strategies that can outperform against a large number of test strategies (opponents) that may not have been seen during coevolution. The second one is concerned with diversity levels in the population that may lead to the search of strategies with poor generalization performance. It is not known if there is a relationship between generalization and diversity in coevolutionary learning. This paper investigates whether there is such a relationship in coevolutionary learning through a detailed empirical study. We systematically investigate the impact of various diversity maintenance approaches on the generalization performance of coevolutionary learning quantitatively using case studies. The problem of the iterated prisoner's dilemma (IPD) game is considered. Unlike past studies, we can measure both the generalization performance and the diversity level of the population of evolved strategies. Results from our case studies show that the introduction and maintenance of diversity do not necessarily lead to the coevolutionary learning of strategies with high generalization performance. However, if individual strategies can be combined (e.g., using a gating mechanism), there is the potential of exploiting diversity in coevolutionary learning to improve generalization performance. Specifically, when the introduction and maintenance of diversity lead to a speciated population during coevolution, where each specialist strategy is capable of outperforming different opponents, the population as a whole can have a significantly higher generalization performance compared to individual strategies.

show abstract

Particle Swarm Optimization Approaches to Coevolve Strategies for the Iterated Prisoner's Dilemma

Cited by 103 publications

References 24 publications

Economía Computacional Basada en Agentes

Economía Computacional Basada en Agentes

Potential AI Strategies to Solve the Commons Game: A Position Paper

Relationship Between Generalization and Diversity in Coevolutionary Learning

Contact Info

Product

Resources

About