HADOUX Emmanuel
责任导师 : Nicolas MAUDET
助理责任导师 : BEYNIER Aurélie, WENG Paul
Markovian sequential decision-making in non-stationary environments: application to argumentative debates
In sequential decision-making problems under uncertainty, an agent makes decisions, one after another, considering the current state of the environment where she evolves. In most work, the environment the agent evolves in is assumed to be stationary, i.e., its dynamics do not change over time. However, the stationarity hypothesis can be invalid if, for instance, exogenous events can occur. In this document, we are interested in sequential decision-making in non-stationary environments. We propose a new model named HS3MDP, allowing us to represent non-stationary problems whose dynamics evolve among a finite set of contexts. In order to efficiently solve those problems, we adapt the POMCP algorithm to HS3MDPs. We also present RLCD with SCD, a new method to learn the dynamics of the environments, without knowing a priori the number of contexts. We then explore the field of argumentation problems, where few works consider sequential decision-making. We address two types of problems: stochastic debates (APS) and mediation problems with non-stationary agents (DMP). In this work, we present a model formalizing APS and allowing us to transform them into an MOMDP in order to optimize the sequence of arguments of one agent in the debate. We then extend this model to DMPs to allow a mediator to strategically organize speak-turns in a debate.
答辩 : 2015-11-26
评委会 :
M. Yann Chevaleyre, LIPN, Univ Paris 13 [Rapporteur]
M. Pierre Marquis, CRIL, Univ Artois [Rapporteur]
Mme. Leila Amgoud (Examinatrice) IRIT
M. Olivier Buffet, LORIA/INRIA
M. Patrice Perny (Examinateur) LIP6, Univ Paris 6
M. Nicolas Maudet, LIP6, Univ Paris 6
Mme. Aurélie Beynier, LIP6, Univ Paris 6
M. Paul Weng, SYSU-CMU, Carnegie Mellon
M. Anthony Hunter (Invité) UCL
2013-2018 刊物
-
2018
- E. Hadoux, A. Beynier, N. Maudet, P. Weng : “Mediation of Debates with Dynamic Argumentative Behaviors”, Computational Models of Argument, vol. 305, Frontiers in Artificial Intelligence and Applications, Warsaw, Poland, pp. 249-256 (2018)
-
2015
- E. Hadoux : “Décision séquentielle markovienne en environnements non-stationnaires : application aux débats d?argumentation”, 博士论文, 答辩 2015-11-26, 责任导师 Maudet, Nicolas, 助理责任导师 : Beynier, Aurélie, Weng, Paul (2015)
- E. Hadoux, A. Beynier, N. Maudet, P. Weng, A. Hunter : “Optimization of Probabilistic Argumentation With Markov Decision Models”, International Joint Conference on Artificial Intelligence, Buenos Aires, Argentina (2015)
-
2014
- E. Hadoux, A. Beynier, P. Weng : “Sequential Decision-Making under Non-stationary Environments via Sequential Change-point Detection”, Learning over Multiple Contexts (LMCE), Nancy, France (2014)
- E. Hadoux, A. Beynier, P. Weng : “Solving Hidden-Semi-Markov-Mode Markov Decision Problems”, AAMAS Workshop Adaptative Learning Agents, ALA 2014, Paris, France (2014)
- E. Hadoux, A. Beynier, P. Weng : “Prise de décision séquentielle en environnements incertains et non stationnaires”, ROADEF - 15e congrès annuel de la Société française de recherche opérationnelle et d'aide à la décision, Bordeaux, France (2014)
- E. Hadoux, A. Beynier, P. Weng : “Solving Hidden-Semi-Markov-Mode Markov Decision Problems”, Scalable Uncertainty Management, vol. 8720, Lecture Notes in Computer Science, Oxford, United Kingdom, pp. 176-189, (Springer International Publishing) (2014)
-
2013
- E. Hadoux, A. Beynier, P. Weng : “Apprentissage de politique par minimisation de regret”, 14e Congrès de la Société Française de Recherche Opérationnelle et d'Aide à la Décision (ROADEF 2013), Troyes, France (2013)