nach oben

2011 | Buch

Kapitel lesen Erstes Kapitel lesen

Multi-Agent-Based Simulation XI

International Workshop, MABS 2010, Toronto, Canada, May 11, 2010, Revised Selected Papers

herausgegeben von: Tibor Bosse, Armando Geller, Catholijn M. Jonker

Verlag: Springer Berlin Heidelberg

Buchreihe : Lecture Notes in Computer Science

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

This volume contains a selection of the papers presented at the 11th International Workshop on Multi-Agent-Based Simulation (MABS 2010), a workshop co-located with the 9th International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS 2010), which was held on May 10-14, 2010 in Toronto, Canada. The 11 revised full papers presented were carefully reviewed and selected from 26 submissions. The workshop has been an important source of inspiration for the body of knowledge that has been produced in the field of Multi-Agent Systems (MAS). As illustrated by this volume, the workshop continues to bring together researchers interested in MAS engineering with researchers focused on finding efficient ways to model complex social systems in social, economic and organizational areas. In all these areas, agent theories, metaphors, models, analyses, experimental designs, empirical studies, and methodological principles all converge into simulation as a way of achieving explanations and predictions, exploring and testing hypotheses, and producing better designs and systems.

Inhaltsverzeichnis

Frontmatter

Models and Frameworks for MAS Development

Situational Programming: Agent Behavior Visual Programming for MABS Novices

Abstract

This paper presents an agent-oriented visual programming approach which aims at providing MABS end-users with a means to easily elaborate artificial autonomous behaviors according to a targeted domain, namely situational programming (SP). More specifically, SP defines design principles which could be used to develop MABS visual programming toolkits suited for non developers and MABS novices. This paper presents SP and how it is used to build a MABS video game which can be played by MABS novices, that is any Internet user.

Fabien Michel, Jacques Ferber, Pierre-Alain Laur, Florian Aleman

IRM4MLS: The Influence Reaction Model for Multi-Level Simulation

Abstract

In this paper, a meta-model called IRM4MLS, that aims to be a generic ground to specify and execute multi-level agent-based models is presented. It relies on the influence/reaction principle and more specifically on IRM4S [13,14]. Simulation models for IRM4MLS are defined. The capabilities and possible extensions of the meta-model are discussed.

Gildas Morvan, Alexandre Veremme, Daniel Dupont

Toward a Myers-Briggs Type Indicator Model of Agent Behavior in Multiagent Teams

Abstract

This paper explores the use of the Myers-Briggs Type Indicator (MBTI) as the basis for defining the personality of an agent. The MBTI is a well-known psychological theory of human personality. In the MBTI model, four axes are defined to explain how humans perceive their environment, how they interact with others and how they make decisions based on these traits. The work described here presents a preliminary model of agent behavior in which two of the axes are implemented, combining to reflect four distinct agent personality types. Experiments were conducted under three environmental conditions: single agent setting, homogeneous multiagent team, and heterogeneous multiagent team. Results are presented for each condition and are analyzed in comparison with the other conditions, as well as within the context of the expected MBTI behaviors given each environment and the simulated task. It is demonstrated that agents of each personality type produce very different results, distinct for and characteristic of each MBTI personality type.

Jordan Salvit, Elizabeth Sklar

Exploring MAS Behaviors

Pheromones, Probabilities, and Multiple Futures

Abstract

Most agent-based modeling techniques generate only a single trajectory in each run, greatly undersampling the space of possible trajectories. Swarming agents can explore many alternative futures in parallel, particularly when they interact through digital pheromone fields. This paper shows how these fields and other artifacts developed by such a model can be interpreted as conditional probabilities estimated by sampling a very large number of possible trajectories. This interpretation offers several benefits. It supports theoretical insight into the behavior of swarming models by mapping them onto more traditional probabilistic models such as Markov decision processes, it allows us to derive more information from them than swarming models usually yield, and it facilitates integrating them with probability-based AI mechanisms such as HMM’s or Bayesian networks.

H. Van Dyke Parunak

Finding Forms of Flocking: Evolutionary Search in ABM Parameter-Spaces

Abstract

While agent-based models (ABMs) are becoming increasingly popular for simulating complex and emergent phenomena in many fields, understanding and analyzing ABMs poses considerable challenges. ABM behavior often depends on many model parameters, and the task of exploring a model’s parameter space and discovering the impact of different parameter settings can be difficult and time-consuming. Exhaustively running the model with all combinations of parameter settings is generally infeasible, but judging behavior by varying one parameter at a time risks overlooking complex nonlinear interactions between parameters. Alternatively, we present a case study in computer-aided model exploration, demonstrating how evolutionary search algorithms can be used to probe for several qualitative behaviors (convergence, non-convergence, volatility, and the formation of vee shapes) in two different flocking models. We also introduce a new software tool (BehaviorSearch) for performing parameter search on ABMs created in the NetLogo modeling environment.

Forrest Stonedahl, Uri Wilensky

Game Theory and Information Sharing

On the Profitability of Incompetence

Abstract

The exchange of information is in many multi-agent systems the essential form of interaction. For this reason, it is crucial to keep agents from providing unreliable information. However, agents that provide information have to balance between being highly competent, in order to achieve a good reputation as information provider, and staying incompetent, in order to minimize the costs of information acquisition. In this paper, we use a multi-agent simulation to identify conditions under which it is profitable for agents either to make an investment to become competent, or to economize and stay incompetent. We focus on the case where the quality of the acquired information cannot objectively be assessed in any immediate way and where hence the information end users have to rely on secondary methods for assessing the quality of the information itself, as well as the trustworthiness of those who provide it.

Eugen Staab, Martin Caminada

Mechanisms for the Self-organization of Peer Groups in Agent Societies

Abstract

New mechanisms for group self-organization in agent societies are investigated and examined in the context of sharing digital goods. Specifically we illustrate how cooperative sharers and uncooperative free riders can be placed in different groups of an electronic society in a decentralized manner. We have simulated a decentralized, open P2P system which self-organizes itself to avoid cooperative sharers being exploited by uncooperative free riders. Inspired by human society, we use social mechanisms such as tags, gossip and ostracism. This approach encourages sharers to move to better groups and restricts free riders without necessitating centralized control, which makes the system appropriate for current open P2P systems.

Sharmila Savarimuthu, Maryam Purvis, Martin Purvis, Bastin Tony Roy Savarimuthu

Multigame Dynamics: Structures and Strategies

Abstract

The dominant strategy among game theorists is to pose a problem narrowly, formalize that structure, and then pursue analytical solutions. This strategy has achieved a number of stylized insights, but has not produced nuanced game-theoretic solutions to larger and more complex issues such as extended international historical conflicts, or the detailed assessment of variegated policy alternatives. In order to model more complex historical and policy-oriented processes, it has been proposed that a broader computational approach to game theory that has the potential to capture richer forms of social dynamics be used, namely the ‘multigame.’ In the multigame approach there are multiple games each of which is open, prototypical, implicit, reciprocal, positional, variegated and historical. When later implemented, the multigame approach will offer the potential to rigorously model complex international historical conflicts and variegated policy alternatives that, heretofore, typically required qualitative analysis.

David L. Sallach, Michael J. North, Eric Tatara

MAS in Economics and Negotiation

Microstructure Dynamics and Agent-Based Financial Markets

Abstract

One of the essential features of the agent-based financial models is to show how price dynamics is affected by the evolving microstructure. Empirical work on this microstructure dynamics is, however, built upon highly simplified and unrealistic behavioral models of financial agents. Using genetic programming as a rule-inference engine and self-organizing maps as a clustering machine, we are able to reconstruct the possible underlying microstructure dynamics corresponding to the underlying asset. In light of the agent-based financial models, we further examine the microstructure both in terms of its short-term dynamics and long-term distribution. The time series of the TAIEX is employed as an illustration of the implementation of the idea.

Shu-Heng Chen, Michael Kampouridis, Edward Tsang

Computational Modeling of Culture’s Consequences

Abstract

This paper presents an approach to formalize the influence of culture on the decision functions of agents in social simulations. The key components are (a) a definition of the domain of study in the form of a decision model, (b) knowledge acquisition based on a dimensional theory of culture, resulting in expert validated computational models of the influence of single dimensions, and (c) a technique for integrating the knowledge about individual dimensions. The approach is developed in a line of research that studies the influence of culture on trade processes. Trade is an excellent subject for this study of culture’s consequences because it is ubiquitous, relevant both socially and economically, and often increasingly cross-cultural in a globalized world.

Gert Jan Hofstede, Catholijn M. Jonker, Tim Verwaart

Agent-Based Simulation Modelling of Housing Choice and Urban Regeneration Policy

Abstract

Phenomena in the housing market can be recreated and analysed using the technique of agent-based modelling. Housing policies introduced as part of urban regeneration often seek to address problems of deprivation in segregated communities by introducing the concept of mixed communities, that is, communities mixed by housing tenure and housing type. In this paper, a framework for the creation of a model of housing choice and regeneration policy is presented.

René Jordan, Mark Birkin, Andrew Evans

Backmatter

Titel: Multi-Agent-Based Simulation XI
herausgegeben von: Tibor Bosse
Armando Geller
Catholijn M. Jonker
Verlag: Springer Berlin Heidelberg
Electronic ISBN: 978-3-642-18345-4
Print ISBN: 978-3-642-18344-7
DOI: https://doi.org/10.1007/978-3-642-18345-4