Ons Nefla, PhD

In charge of lecture, tutorials, and lab of Python, C++ and Database.

The thesis project is part of a collaboration between emlyon business school and Paris-Dauphine university. The subject is: Agent-based Group Decision Making for Ordinal Classification : preference elicitation and negociation-aggregation. During my thesis, I develop an ordinal classification process for group decision aiding. For this aim, I have developed a two-phase process. The first pre-negotiation phase consists in assisting a single decision maker in her multi-criteria ordinal classification problem based on a new "incremental" elicitation process. This process selects the most informative questions to ask the decision maker in order to learn her preferences. The second negotiation phase aims at converging several decision makers towards a group classification based on their individual classifications (derived from the pre-negotiation phase). The process is based on a multi agent system to represent the human decision makers and toconduct the process in a distributed and autonomous way. This process aims at satisfying fairly all the decision makers involved by taking into account the evolution of their behaviors and preferences and by optimizing the disclosure of their individual classifications.