Didier Dubois

Centre National de la Recherche Scientifique, Université Toulouse III - Paul Sabatier

Papers

Total Citations

130

H-Index

About

Didier Dubois is a pioneering figure in the fields of fuzzy logic, possibility theory, and artificial intelligence, with a career dedicated to formalizing uncertainty and imprecision in reasoning systems. His most influential work, "On the Combination of Evidence in Various Mathematical Frameworks" (1992), which has garnered over 125 citations, provides a foundational synthesis of how evidence can be merged across probabilistic, possibilistic, and belief-function models—a critical contribution to decision-making under uncertainty. Dubois is also known for advancing qualitative possibilistic planning, as seen in his work on "Qualitative Possibilistic Mixed-Observable MDPs" (2013), where he extended Markov decision processes to handle imprecise beliefs and observations, offering robust solutions for AI systems with limited data. Additionally, his early research on "Shape Understanding via Fuzzy Models" (1986) explored how fuzzy set theory can interpret geometric forms, showcasing his versatility. A co-author of seminal texts on possibility theory, Dubois’s impact spans over four decades, influencing fields from robotics to risk analysis. His work remains essential reading for students and researchers tackling uncertainty in complex systems.