Suchergebnisse

International audience ; In this paper, we introduce a new agent-based method to build a decision-aid tool aimed to improve policy design. In our approach, a policy is defined as a set of levers, modelling the set of actions, the means to impact a complex system. Our method is generic, as it could be applied to any domain, and be coupled with any agent-based simulator. We could deal not only with simple levers (a single variable whose value is modified) but also complex ones (multiple variable modifications, qualitative effects, .), unlike most optimization methods. It is based on the evolutionary algorithm CMA-ES, coupled with a normalized and aggregated fitness function. The fitness is normalized using estimated Ideal (best policy) and Nadir (worst policy) values, these values being dynamically computed during the execution of CMA-ES through a Pareto Front estimated with the ABM simulation. Moreover , to deal with complex levers, we introduce the FSM-branching algorithm, where a Finite State Machine (FSM) determines whether a complex policy can potentially be improved or has to be aborted. We tested our method with Economic Policies on the French Labor Market (FLM), allowing the modification of multiple elements of the FLM, and we compared the results to the reference, the FLM without any policy applied. The policies studied here comprise simple and complex levers. This experience shows the viability of our approach, the efficiency of our algorithms and illustrates how this combination of evolutionary optimization, multi-criteria aggregation and agent-based simulation could help any policy-maker to design better policies.

International audience ; In this paper, we introduce a new agent-based method to build a decision-aid tool aimed to improve policy design. In our approach, a policy is defined as a set of levers, modelling the set of actions, the means to impact a complex system. Our method is generic, as it could be applied to any domain, and be coupled with any agent-based simulator. We could deal not only with simple levers (a single variable whose value is modified) but also complex ones (multiple variable modifications, qualitative effects, .), unlike most optimization methods. It is based on the evolutionary algorithm CMA-ES, coupled with a normalized and aggregated fitness function. The fitness is normalized using estimated Ideal (best policy) and Nadir (worst policy) values, these values being dynamically computed during the execution of CMA-ES through a Pareto Front estimated with the ABM simulation. Moreover , to deal with complex levers, we introduce the FSM-branching algorithm, where a Finite State Machine (FSM) determines whether a complex policy can potentially be improved or has to be aborted. We tested our method with Economic Policies on the French Labor Market (FLM), allowing the modification of multiple elements of the FLM, and we compared the results to the reference, the FLM without any policy applied. The policies studied here comprise simple and complex levers. This experience shows the viability of our approach, the efficiency of our algorithms and illustrates how this combination of evolutionary optimization, multi-criteria aggregation and agent-based simulation could help any policy-maker to design better policies.

Cet article présente un modèle multi-agent du marché du travail français et l'utilise pour une simulation ex ante des effets du contrat de génération. Le modèle simule les décisions et les interactions des individus et des entreprises (10 000 agents dans nos simulations), dans le cadre d'un environnement économique et juridique spécifié. Les décisions consistent pour les entreprises à créer ou supprimer des emplois de type CDD et CDI, ainsi qu'à embaucher, promouvoir et licencier leurs salariés. Les individus candidatent aux emplois et peuvent démissionner. Les interactions engendrent les flux bruts entre les états de salarié, chômeur et inactif, formant une comptabilité stock-flux complète. Un algorithme de calibrage permet de reproduire les principales caractéristiques du marché du travail français en 2011, fortement structuré par la coexistence des CDI et des CDD. Tel que simulé par le modèle, le contrat de génération a bien diminué le chômage des jeunes, mais cet effet est modeste en raison d'effets d'aubaine très importants ainsi que d'effets d'éviction modérés. Les trajectoires simulées des jeunes qui ont bénéficié du contrat de génération montrent un effet positif sur leur taux d'emploi à long terme. Des variantes montrent qu'il n'est pas aisé d'améliorer le dispositif.

International audience ; Cet article présente un modèle multi-agents du marché du travail français et l'utilise pour analyser les eets du contrat de génération, simulés ex-ante. Notre approche consiste à produire un modèle computationnel des décisions et des interactions des individus et des entreprises (10 000 agents dans nos simulations), dans le cadre d'un environnement économique et juridique spécié. Les décisions consistent pour les entreprises à créer ou supprimer des emplois de type CDD et CDI, en réponse à des chocs individuels, ainsi qu'à embaucher, promouvoir et licencier leurs salariés. Les individus candidatent aux emplois, et peuvent démissionner, soit pour un autre emploi, soit pour abandonner l'activité. Les interactions engendrent les ux bruts entre les états de salarié, chômeur et inactif, formant une comptabilité stock-ux complète. Un algorithme de calibrage permet de reproduire les principales caractéristiques du marché du travail français en 2011. La coexistence de CDI et de CDD structure profondément le fonctionnement du marché, car si des jeunes utilisent les CDD pour s'insérer dans des CDI, d'autres alternent entre CDD et chômage, et cette mobilité subsiste à l'âge adulte. Le Contrat de Génération (CG) vise à faciliter la création de CDI pour les jeunes ainsi qu'à développer leurs compétences par la transmission de savoir de la part de seniors dont le maintien dans l'emploi est simultanément encouragé. L'eet d'insertion des jeunes est vérié dans le modèle, mais cet eet est fort modeste du fait d'eets d'aubaine très importants ainsi que d'eets d'éviction modérés. Le coût par chômeur évité apparaît inférieur à celui des allègements de charge. L'article étudie enn les trajectoires des jeunes qui ont bénécié du CG et montre un eet positif sur un terme long. Des variantes montrent qu'il n'est pas aisé d'améliorer le dispositif. Les résultats chirés doivent naturellement être considérés avec prudence et le modèle vise d'abord à clarier l'analyse des nombreuses interactions en ÷uvre.

International audience ; Cet article présente un modèle multi-agents du marché du travail français et l'utilise pour analyser les eets du contrat de génération, simulés ex-ante. Notre approche consiste à produire un modèle computationnel des décisions et des interactions des individus et des entreprises (10 000 agents dans nos simulations), dans le cadre d'un environnement économique et juridique spécié. Les décisions consistent pour les entreprises à créer ou supprimer des emplois de type CDD et CDI, en réponse à des chocs individuels, ainsi qu'à embaucher, promouvoir et licencier leurs salariés. Les individus candidatent aux emplois, et peuvent démissionner, soit pour un autre emploi, soit pour abandonner l'activité. Les interactions engendrent les ux bruts entre les états de salarié, chômeur et inactif, formant une comptabilité stock-ux complète. Un algorithme de calibrage permet de reproduire les principales caractéristiques du marché du travail français en 2011. La coexistence de CDI et de CDD structure profondément le fonctionnement du marché, car si des jeunes utilisent les CDD pour s'insérer dans des CDI, d'autres alternent entre CDD et chômage, et cette mobilité subsiste à l'âge adulte. Le Contrat de Génération (CG) vise à faciliter la création de CDI pour les jeunes ainsi qu'à développer leurs compétences par la transmission de savoir de la part de seniors dont le maintien dans l'emploi est simultanément encouragé. L'eet d'insertion des jeunes est vérié dans le modèle, mais cet eet est fort modeste du fait d'eets d'aubaine très importants ainsi que d'eets d'éviction modérés. Le coût par chômeur évité apparaît inférieur à celui des allègements de charge. L'article étudie enn les trajectoires des jeunes qui ont bénécié du CG et montre un eet positif sur un terme long. Des variantes montrent qu'il n'est pas aisé d'améliorer le dispositif. Les résultats chirés doivent naturellement être considérés avec prudence et le modèle vise d'abord à clarier l'analyse des nombreuses interactions en ÷uvre.

International audience ; Attitude is a key concept in social psychology. The paper presents a novel agent-based model to simulate attitude formation by combining a rational and an emotional components based on cognitive, psychological and social theories. Individuals of the artificial population perceive actions taken by actors such as government or brands, they form an attitude toward them and also communicate the events through a social network. The model outputs are first studied through a functional analysis in which some unique macroscopic behaviors have emerged such as the impact of social groups, the resistance of the population toward disinfor-mation campaigns or the social pressure. We then applied our model on a real world scenario depicting the eeort of French Forces in their stabilization operations in Kapisa (Afghanistan) between and. We calibrated the model parameters based on this scenario and the results of opinion polls that were conducted in the area during the same period about the sentiment of the population toward the Forces. Our model was able to reproduce polls results with a global error under %. Based on these results, we show the diierent dynamics tendencies that emerged among the population by applying a non-supervised classification algorithm.

International audience ; Attitude is a key concept in social psychology. The paper presents a novel agent-based model to simulate attitude formation by combining a rational and an emotional components based on cognitive, psychological and social theories. Individuals of the artificial population perceive actions taken by actors such as government or brands, they form an attitude toward them and also communicate the events through a social network. The model outputs are first studied through a functional analysis in which some unique macroscopic behaviors have emerged such as the impact of social groups, the resistance of the population toward disinfor-mation campaigns or the social pressure. We then applied our model on a real world scenario depicting the eeort of French Forces in their stabilization operations in Kapisa (Afghanistan) between and. We calibrated the model parameters based on this scenario and the results of opinion polls that were conducted in the area during the same period about the sentiment of the population toward the Forces. Our model was able to reproduce polls results with a global error under %. Based on these results, we show the diierent dynamics tendencies that emerged among the population by applying a non-supervised classification algorithm.

International audience ; Attitude is a key concept in social psychology. The paper presents a novel agent-based model to simulate attitude formation by combining a rational and an emotional components based on cognitive, psychological and social theories. Individuals of the artificial population perceive actions taken by actors such as government or brands, they form an attitude toward them and also communicate the events through a social network. The model outputs are first studied through a functional analysis in which some unique macroscopic behaviors have emerged such as the impact of social groups, the resistance of the population toward disinfor-mation campaigns or the social pressure. We then applied our model on a real world scenario depicting the eeort of French Forces in their stabilization operations in Kapisa (Afghanistan) between and. We calibrated the model parameters based on this scenario and the results of opinion polls that were conducted in the area during the same period about the sentiment of the population toward the Forces. Our model was able to reproduce polls results with a global error under %. Based on these results, we show the diierent dynamics tendencies that emerged among the population by applying a non-supervised classification algorithm.

International audience ; This paper presents a multi-agent model for simulating attitude formation and change based on perception and communication in the context of stabilization operations. The originality of our model comes from (1) attitude computation that evaluates information as part of a history relative to the individual (2) a notion of co-responsibility for attitude attribution. We present a military scenario of French operations in Afghanistan along with polls results about the opinion of citizen toward present Forces. Based on these field data, we calibrate the model and show the resulting attitude dynamics. We study the sensibility of the model to the co-responsibility factor.

International audience ; This paper presents a multi-agent model for simulating attitude formation and change based on perception and communication in the context of stabilization operations. The originality of our model comes from (1) attitude computation that evaluates information as part of a history relative to the individual (2) a notion of co-responsibility for attitude attribution. We present a military scenario of French operations in Afghanistan along with polls results about the opinion of citizen toward present Forces. Based on these field data, we calibrate the model and show the resulting attitude dynamics. We study the sensibility of the model to the co-responsibility factor.

International audience ; This paper presents a multi-agent model for simulating attitude formation and change based on perception and communication in the context of stabilization operations. The originality of our model comes from (1) attitude computation that evaluates information as part of a history relative to the individual (2) a notion of co-responsibility forattitude attribution. We present a military scenario of French operations in Afghanistan along with polls results about the opinion of citizen toward present Forces. Based on these eld data, we calibrate the model and show the resulting attitude dynamics. We study the sensibility of the model to the co-responsibility factor.

International audience ; This paper presents a multi-agent model for simulating attitude formation and change based on perception and communication in the context of stabilization operations. The originality of our model comes from (1) attitude computation that evaluates information as part of a history relative to the individual (2) a notion of co-responsibility for attitude attribution. We present a military scenario of French operations in Afghanistan along with polls results about the opinion of citizen toward present Forces. Based on these field data, we calibrate the model and show the resulting attitude dynamics. We study the sensibility of the model to the co-responsibility factor.

International audience ; This paper presents a multi-agent model for simulating attitude formation and change based on perception and communication in the context of stabilization operations. The originality of our model comes from (1) attitude computation that evaluates information as part of a history relative to the individual (2) a notion of co-responsibility forattitude attribution. We present a military scenario of French operations in Afghanistan along with polls results about the opinion of citizen toward present Forces. Based on these eld data, we calibrate the model and show the resulting attitude dynamics. We study the sensibility of the model to the co-responsibility factor.

International audience ; This paper presents a multi-agent model for simulating attitude formation and change based on perception and communication in the context of stabilization operations. The originality of our model comes from (1) attitude computation that evaluates information as part of a history relative to the individual (2) a notion of co-responsibility forattitude attribution. We present a military scenario of French operations in Afghanistan along with polls results about the opinion of citizen toward present Forces. Based on these eld data, we calibrate the model and show the resulting attitude dynamics. We study the sensibility of the model to the co-responsibility factor.

International audience ; In the context of military training for stabilizationoperation of a crisis zone with civilian population, understandingthe formation of attitude and its dynamics is a key issue. Thispaper presents a multi-agent model for simulating attitude forma-tion and change based on individuals perception of informationand its diffusion through communication. We represent theattitude as object-evaluation associations of varying strength pro-posed by Fazio [1]. Individuals observe military operations. Theyexchange and revise beliefs about social objects depending onmultiple criteria deriving from social psychology theories. Theycompute their attitude value based on analytic assessment of thesebeliefs. We illustrate, through several simulation experiments, therole of communication on attitude dynamics.