Academic literature on the topic 'Rationalisabilité'

Different agents may have different points of view. Following a popular approach in the artificial intelligence literature, this can be modeled by means of different abstract argumentation frameworks, each consisting of a set of arguments the agent is contemplating and a binary attack-relation between them. A question arising in this context is whether the diversity of views observed in such a profile of argumentation frameworks is consistent with the assumption that every individual argumentation framework is induced by a combination of, first, some basic factual attack-relation between the arguments and, second, the personal preferences of the agent concerned regarding the moral or social values the arguments under scrutiny relate to. We treat this question of rationalisability of a profile as an algorithmic problem and identify tractable and intractable cases. In doing so, we distinguish different constraints on admissible rationalisations, e.g., concerning the types of preferences used or the number of distinct values involved. We also distinguish two different semantics for rationalisability, which differ in the assumptions made on how agents treat attacks between arguments they do not report. This research agenda, bringing together ideas from abstract argumentation and social choice, is useful for understanding what types of profiles can reasonably be expected to occur in a multiagent system.

Les préférences révélées lient choix, préférences et bien-être lorsque les choix apparaissent cohérents. Le premier chapitre s’intéresse à la force des hypothèses nécessaire pour obtenir des indications précises sur le bien-être quand les choix sont incohérents. Il utilise les données d’expériences en laboratoire et sur le terrain pour évaluer le pouvoir prédictif de deux approches utilisant peu d’hypothèses. Ces approches ont un pouvoir prédictif élevé pour une majorité d’individus, elles fournissent donc des indications précises sur le bien-être. Le pouvoir prédictif de ces approches est fortement corrélé à deux propriétés des préférences révélées. Le deuxième chapitre introduit une méthode pour obtenir l’ensemble des meilleures alternatives d’un individu, en cohérence avec théorie des préférences révélées, mais en contradiction avec les pratiques expérimentales. Les individus sont incités à choisir plusieurs alternatives à l’aide d’un petit paiement additionnel, mais sont rémunérés à la in par une seule, tirée au hasard. Les conditions pour que les meilleures alternatives soient partiellement ou complètement identifiées sont données. Le troisième chapitre applique cette méthode dans une expérience. Les meilleures alternatives sont complètement identifiées pour 18% des sujets et partiellement pour 40%. Les préférences complètes, réflexives et transitives rationalisent 40% des choix observés dans l’expérience. Permettre que les choix dépendant de l’ensemble de choix, tout en conservant les préférences classiques, rationalise 96% des choix observés. Enfin, on observe une quantité significative d’indifférence, bien supérieure à ce qui est obtenu traditionnellement
Revealed preferences link choices, preferences, and welfare when choices appear consistent. The first chapter assesses how much structure is necessary to impose on a model to provide precise welfare guidance based on inconsistent choices. We use data sets from the lab and field to evaluate the predictive power of two conservative “model-free” approaches of behavioral welfare analysis. We find that for most individuals, these approaches have high predictive power, which means there is little ambiguity about what should be selected from each choice set. We show that the predictive power of these approaches correlates highly with two properties of revealed preferences. The second chapter introduces a method for eliciting the set of best alternatives of decision makers, in line with the theory on revealed preferences, but at odds with the current practice. We allow decision makers to choose several alternatives, provide an incentive for each alternative chosen, and then randomly select one for payment. We derive the conditions under which we partially or fully identify the set of best alternatives. The third chapter applies the method in an experiment. We fully identify the set of best alternatives for 18% of subjects and partially identify it for another 40%. We show that complete, reflexive, and transitive preferences rationalize 40% of observed choices in the experiment. Going beyond, we show that allowing for menu-dependent choices while keeping classical preferences rationalize 96% of observed choices. Besides, eliciting sets allows us to conclude that indifference is significant in the experiment, and underestimate by the classical method

We review a recently introduced model in which each of a number of agents is endowed with an abstract argumentation framework reflecting her individual views regarding a given set of arguments. A question arising in this context is whether the diversity of views observed in such a situation is consistent with the assumption that every individual argumentation framework is induced by a combination of, first, some basic factual information and, second, the personal preferences of the agent concerned. We treat this question of rationalisability of a profile as an algorithmic problem and identify tractable and intractable cases. This is useful for understanding what types of profiles can reasonably be expected to occur in a multiagent system.