A Model of Co-evolution in Multi-agent System

Dreżewski, Rafał

doi:10.1007/3-540-45023-8_30

Cited by 40 publications

(23 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the other population, neurons received fitness based upon the number of successful blueprints they participated in. Drezewski [11] describes a multi-agent system that uses a cooperative coevolutionary process with two species. The system solves multi-objective optimization tasks via locating the pareto frontier as a result of co-evolutionary interactions between the species.…”

Section: Cooperative Co-evolutionmentioning

confidence: 99%

Collective neuro-evolution for evolving specialized sensor resolutions in a multi-rover task

2009

View full text Add to dashboard Cite

This article presents results from an evaluation of the collective neuro-evolution (CONE) controller design method. CONE solves collective behavior tasks, and increases task performance via facilitating emergent behavioral specialization. Emergent specialization is guided by genotype and behavioral specialization difference metrics that regulate genotype recombination. CONE is comparatively tested and evaluated with similar neuro-evolution methods in an extension of the multi-rover task, where behavioral specialization is known to benefit task performance. The task is for multiple simulated autonomous vehicles (rovers) to maximize the detection of points of interest (red rocks) in a virtual environment. Results indicate that CONE is appropriate for deriving sets of specialized rover behaviors that complement each other such that a higher task performance, comparative to related controller design methods, is attained in the multi-rover task.

show abstract

Section: Cooperative Co-evolutionmentioning

confidence: 99%

Collective neuro-evolution for evolving specialized sensor resolutions in a multi-rover task

2009

View full text Add to dashboard Cite

show abstract

“…A comprehensive review of existing niching techniques has been presented in [6]. They have been divided into three classes of algorithms, based on:…”

Section: Niching In Classical Evolutionary Algorithmsmentioning

confidence: 99%

“…An analysis of such applicability, though restricted to coevolutionary multi-agent algorithms, has been performed in [6]. However, it is still an open question to what extent these techniques can be used in simpler and more general EMAS.…”

Section: Introductionmentioning

confidence: 99%

Niching in Evolutionary Multi-agent Systems

Krzywicki¹

2013

csci

View full text Add to dashboard Cite

Niching is a group of techniques used in evolutionary algorithms, useful in several types of problems, including multimodal or nonstationary optimization. This paper investigates the applicability of these methods to evolutionary multi-agent systems (EMAS), a hybrid model combining the advantages of evolutionary algorithms and multi-agent systems. This could increase the efficiency of this type of algorithms and allow to apply them to a wider class of problems. As a starting point, a simple but flexible EMAS framework is proposed. Then, it is shown how to extend this framework in order to introduce niching, by adapting two classical niching methods. Finally, preliminary experimental results show the efficiency and the simultaneous discovery of multiple optima by this modified EMAS.

show abstract

“…Basic EMAS model also suffers from the negative tendency to loss the population diversity, however, as we will show in the following sections, it can be equipped with additional mechanisms and operators which improve the quality of obtained results. The general model of co-evolutionary multi-agent system (CoEMAS) [6] introduces additionally the notions of species, sexes, and interactions between them. CoEMAS allows modeling and simulation of different co-evolutionary interactions, which can serve as the basis for constructing the techniques of maintaining population diversity and improving adaptive capabilities of such systems (for example see [7]).…”

Section: Introductionmentioning

confidence: 99%

Techniques for Maintaining Population Diversity in Classical and Agent-Based Multi-objective Evolutionary Algorithms

Dreżewski

Siwik

2007

Computational Science – ICCS 2007

Self Cite

View full text Add to dashboard Cite

Abstract. The loss of population diversity is one of the main problems in some applications of evolutionary algorithms. In order to maintain useful population diversity some special techniques must be used, like niching or co-evolutionary mechanisms. In this paper the mechanisms for maintaining population diversity in agent-based multi-objective (co-)evolutionary algorithms are proposed. The presentation of techniques is accompanied by the results of experiments and comparisons to "classical" evolutionary multi-objective algorithms.

show abstract

A Model of Co-evolution in Multi-agent System

Cited by 40 publications

References 4 publications

Collective neuro-evolution for evolving specialized sensor resolutions in a multi-rover task

Collective neuro-evolution for evolving specialized sensor resolutions in a multi-rover task

Niching in Evolutionary Multi-agent Systems

Techniques for Maintaining Population Diversity in Classical and Agent-Based Multi-objective Evolutionary Algorithms

Contact Info

Product

Resources

About