Parameter Space Exploration of Agent-Based Models

Computational Complexity, 2012

H ow is it possible that a whole ancient civilization disappeared? Was this caused by climate changes? What type of recruitment strategy among social insects is best adapted to their particular environment? How much time does it take to evacuate an airport if people have limited perception caused by smoke as well as restricted mobility? What if many of these people travel in groups or families? How long does it take for commuters to reach their destinations if an important arterial in the Los Angeles area is closed? These are the kind of questions that can be and are answered by agent-based modeling and simulation (ABMS). In this paradigm, simulated human beings or animals are modeled as agents, interacting with some of their peers as well as with their environment. The environment, as in many multiagent systems, plays a key role and must therefore be carefully taken into account. For instance, passengers seeking to leave the airport just mentioned try to find the shortest way to an exit, which may be partially hindered by debris. These are only some examples of scenarios-also characterized as complex adaptive systems-that can be investigated using ABMS. The core idea here is to use simulated agents for producing a phenomenon that shall be analyzed, reproduced, or predicted. This generative, bottom-up nature of modeling and simulation provides great potential for dealing with problems in which conventional modeling and simulation paradigms have difficulties capturing the core features of the original system. In what follows, this particular modeling and simulation paradigm, its concept, properties, and application are introduced Articles

Winter Simulation Conference, 2005

Agent-based modelling and simulation (ABMS) is a relatively new approach to modelling systems composed of autonomous, interacting agents. Agent-based modelling is a way to model the dynamics of complex systems and complex adaptive systems. Such systems often self-organize themselves and create emergent order. Agent-based models also include models of behaviour (human or otherwise) and are used to observe the collective effects of agent behaviours and interactions. The development of agent modelling tools, the availability of micro-data, and advances in computation have made possible a growing number of agent-based applications across a variety of domains and disciplines. This article provides a brief introduction to ABMS, illustrates the main concepts and foundations, discusses some recent applications across a variety of disciplines, and identifies methods and toolkits for developing agent models.

Springer eBooks, 2001

Winter Simulation Conference (WSC), …, 2009

OptAgent is a new framework which enhances the practical applications of agent-based models by giving them an optimization and analysis capability. Embedding such an ability within agent-based systems greatly expands their usefulness and applicability. Developed through an Army Research Office SBIR Phase II grant, OptAgent is built on our existing OptQuest engine to create a system that can be used alike with agent-based and other complex simulation models. Important components of the OptAgent system include new predictors based on Markov blankets and satisfiability data mining that speed the search for high quality solutions. OptAgent is structured to assist with the verification, validation, and accreditation process for a new model and can also be used to enhance the analysis accomplished with a more mature model.

Proceedings. IEEE International Joint Symposia on Intelligence and Systems (Cat. No.98EX174), 1998

Determining an optimal solution is almost impossible but trying to improve an existing solution is a way to lead to a better scheduling. We use a Multi-Agent System guided by a Multi-Objective Genetic Algorithm to find a balance point in the respect of a solution of the Pareto front. Of course, this solution isn't the best but allows a multicriteria optimization. By crossover and mutation of agents, according to their fitness function, we improve an existing solution. Therefore, the construction of some system simulating living organisms or social systems, cannot be modelled using a strictly mechanical approach. They are typically adaptive and their behaviour is not regular. The multi-agent system must express radical characters, such as reification of emergence, property of controlled self-reproduction of groups of agents and not linear behaviour.

… (CEC), 2011 IEEE …, 2011

Despite their increasing popularity, agent-based models are hard to test, and so far no established testing technique has been devised for this kind of software applications. Reverse engineering an agent-based model specification from model simulations can help establish a confidence level about the implemented model and in some cases reveal discrepancies between observed and normal or expected behaviour. In this study, a multiobjective optimisation technique based on a simple random search algorithm is deployed to dynamically infer and refine the specification of three agent-based models from their simulations. The multiobjective optimisation technique also incorporates a dynamic invariant detection technique which serves to guide the search towards uncovering new model behaviour that better captures the model specification. The Non-dominated Sorting Genetic Algorithm (NSGA-II) was also deployed to replace the random search algorithm, and the results from both approaches were compared. While both algorithms revealed good potential in capturing the model specifications, the pure exploratory nature of random search was found more suitable for the application at hand, compared to the balanced exploitation/exploration nature of genetic algorithms in general.

International Journal of Computer Applications, 2015

This paper describes the organizational modeling of the Ant Colony Optimization (ACO). It presents a modeling approach of the ACO based on Holonic Multi-Agent paradigm named HMAS (Holonic Multi-Agent Systems). The approach of modeling used is organizational and it uses four basic concepts: Capacity, Role, Interaction and Organization (CRIO). The Traditional modeling techniques fail to capture interactions between loosely coupled aspects of a complex system. However, the organizational model of the ACO has highlighted the different roles that can occur in such optimization device. The solving approach highlights two fundamental concepts from behavioral intensification and diversification. Since, it is difficult to distinguish an intensification from a diversification behavior, though these two trends are identifiable in the organizational model of the proposed ACO, a single role can combine the roles Intensif y and Diversif y. So, a M anager role is identified and is responsible for the coordination of research by the colonies, and the management of the pheromone memory.

Proceedings of the 2002 ACM symposium on …, 2002

This paper describes some extensions added to the continuous simulation language OOCSMP to perform agent-oriented simulation. The extensions are tested by simulating the evolution of a colony of virtual ants (vants). In this simulation, each vant is modelled as an agent and is assigned a set of genes that control some aspects of its behaviour, such as its velocity, memory, communication abilities, scepticism, etc. Some emergent properties of the swarm of vants have been observed.

Bulletin of Mathematical Biology, 2016

Agent-based models (ABMs) have become an increasingly important mode of inquiry for the life sciences. They are particularly valuable for systems that are not understood well enough to build an equation-based model. These advantages, however, are counterbalanced by the difficulty of analyzing and using ABMs, due to the lack of the type of mathematical tools available for more traditional models, G. An, B. G. Fitzpatrick co-first authors.

2001

In the biological literature on animal behaviour, in addition to real experiments and field studies, also simulation experiments are a useful source of progress. Often specific mathematical modelling techniques are adopted and directly implemented in a programming language. Modelling more complex agent behaviours is less adequate using the usually adopted mathematical modelling techniques. The literature on AI and Agent Technology offers more specific methods to design and implement (also more complex) intelligent agents and agent societies on a conceptual level. One of these methods is the compositional multi-agent system design method DESIRE. In this paper it is shown how (depending on the complexity of the required behaviour) a simulation model for animal behaviour can be designed at a conceptual level in an agent-based manner. Different models are shown for different types of behaviour, varying from purely reactive behaviour to pro-active, social and adaptive behaviour. The compositional design method for multi-agent systems DESIRE and its software environment supports the conceptual and detailed design, and execution of these models. A number of experiments reported in the literature on animal behaviour have been simulated for different agent models.