Problem modelling and solving using constraints
Many tasks in artificial intelligence, and more generally in computer science, require the resolution of difficult problems from a computational point of view.
The “Constraints” axis deals with the modeling and practical solution of of combinatorial problems under constraints. Two formalisms are mainly studied in the axis: SAT (Satisfiability Testing), where the variables are boolean and CSP (Constraint Satisfaction Problem) where the variables are defined on finite domains. These two formalisms are part of the general field of constraint programming (CP for Constraint Programming). Among the problems studied are satisfaction problems (existence of a solution), optimization problems (search for a “best” solution), counting problems (calculation of the number of solutions), enumeration problems (of solutions). The work carried out in this area ranges from theoretical developments (study of complexity, analysis of the properties of languages and algorithms) to the development of software (called solvers) allowing the practical solution of the problems studied.
Among the research themes of the “Constraints” axis, we find :
- the development of general modeling languages and associated tools
- the design of efficient mechanisms for traversing the search space
- the design of inference and learning mechanisms within solvers
- the upstream compilation of combinatorial problems in order to carry out, a posteriori, a set of queries on these problems in an efficient way.
Within the “Constraints” axis, the CRIL is therefore interested in various questions ranging from representation to the resolution of combinatorial problems. Work focuses on the extension of existing frameworks and the use of techniques from SAT, CSP and related formalisms (pseudo-boolean, quantified boolean formulas) to enlarge the fields of application. For example, many of the tools developed in the axis are used to address fundamental problems for explainable and robust AI.
The development of generic open-source solvers and the resolution of real (industrial) problems are also part of the activities conducted within the axis.
Members in the axis
Audemard Gilles | Professor | iut |
Bellart Steve | PhD student | fac |
Blomme Anthony | PhD student | fac |
Bounia Louenas | PhD student | fac |
Bouraoui Zied | Associate professor, accredited to supervise research | fac |
Boussemart Frédéric | Associate professor | iut |
Capelli Florent | Tenure track | |
Chmeiss Assef | Associate professor | iut |
Condotta Jean-François | Professor | iut |
Crowley Daniel | PhD student | |
de Lima Tiago | Associate professor, accredited to supervise research | fac |
Delorme Fabien | Engineer | fac |
D'halluin Cyrille | PAST | fac |
Falque Thibault | PhD student | fac |
Grégoire Éric | Professor | fac |
Hemery Fred | Associate professor | iut |
Ing David | PhD student | fac |
Jabbour Saïd | Professor | fac |
Kemgue Alain Trésor | Engineer | fac |
Koriche Frédéric | Professor | iut |
Lagniez Jean-Marie | Professor | fac |
Le Berre Daniel | Professor | fac |
Lecoutre Christophe | Professor | iut |
Lonca Emmanuel | Engineer | fac |
Marquis Pierre | Professor | fac |
Mazure Bertrand | Professor | fac |
Mengel Stefan | CNRS researcher, accredited to supervise research | fac |
Miceli Marie | PhD student | fac |
Paris Nicolas | PAST | iut |
Parrain Anne | Associate professor | fac |
Piette Cédric | Associate professor | iut |
Roussel Olivier | Associate professor | iut |
Saïs Lakhdar | Professor | fac |
Salhi Yakoub | Professor | iut |
Samy Modeliar Mouny | PRAG | iut |
Tabary Sébastien | Associate professor | iut |
Tahil Gökhan | PhD student | fac |
Wallon Romain | Associate professor | iut |
Ongoing PhDs
- Determination of the Association Constant between a Modified Cyclodextrin and a Guest by Artificial Intelligence for Valorization of Bio-Sourced Substrates by using Green Chemistry Principles - Gökhan Tahil
- Découverte de connaissances à partir de données migratoires - David Ing
- Explanation via the « weak cores » - Marie Miceli
- Formal models for explainable and robust AI - Louenas Bounia
- Knowledge Compilation for explainable and robust AI - Steve Bellart
- Optimal and sub-optimal solutions explanation in Boolean setting - Anthony Blomme
- Optimization of passenger flows and resource management through machine learning and constraint programming techniques - Thibault Falque
- SAT-based Approaches for Formal Verification with B method - Daniel Crowley