CoDE Research CoDE Research
IRIDIA Research IRIDIA Research
SMG Research
WIT Research
WIT Research
SMG Research
Home People Research Activities Publications Teaching Resources
Preference Modeling Robustness Multicriteria Analysis Operational Research
Preference Modeling Robustness Multicriteria Analysis Operational Research
login

Multicriteria Analysis


Multicriteria decision analysis has known a very rapid development during the last twenty years: this is characterized by numerous methods and applications, but by a lack of theoretical bases. Our goal is to improve the theoretical aspect by a systematic study of the properties of the methods and by the elaboration of axiomatic foundations to classify the methods. We have also proposed new methods or improved some existing ones, on basis of the obtained theoretical results.
  1. Theoretical contributions
    A theoretical approach which was adopted consists in defining, a priori, a set of properties that may be considered as desirable for a preference aggregation method. The existing methods are then sorted in function of these properties and new methods can be proposed. Of course, the set of properties may vary according to the type of decision problem which is considered. The result of this approach is a methodology allowing to choose aggregation procedures resting on solid theoretical foundations, to justify the use of a procedure in function of the type of problem and to be able to explain and interpret the obtained results on basis of the properties of the methods. This research is partially connected to social choice theory.
  2. Methodological contributions
    The methodological contributions are mainly concerned with multiobjective programming and outranking methods. An active research direction is devoted to the multicriteria classification problems. Among them, we can mention the ordered clustering problem or the management of disjunctive categories.
  3. Group MCDA in system dynamics
    The dynamic control of human systems is central to many decision problems: for example, in the definition of macro-economic and environmental policies by governments or in the management of private companies. A control framework has been developed as a prospective tool also usable for real-time control. This methodology combines Group MCDA with System Dynamics (SD). It proceeds in three stages. In the first stage a SD model is implemented for the purpose of simulations. In the second stage, control policies are selected using Group MCDA on the basis of these simulations. In the third stage, control is exerted, either in real-time in case of actual policy implementation, or in virtual future scenarios in case of prospective investigations. Control policies are permanently monitored and adapted using Group MCDA.
  4. Applications
    Members of the SMG have been involved in a number of real-world applications such as in environmental management, finance, medical diagnosis, energy production and evaluation problematic, software development, or nuclear emergency strategy.

Projects

List of the relevant projects on the thematic area:

  • EVAL: Evaluation Methodology and Secure Generic Software in a multi-users environment
  • iVISION: Développement d'un logiciel de visualisation pour l'aide multicritère à la décision

Publications

List of publications concerning this thematic area:


Updated: 2017-03-27