Machine Learning for Optimization
Solving hard optimization problems in real-world applications using data-driven approaches, combining learning and optimization.
Optimization problems are ubiquitous in many fields such as robotics, finance, logistics, transportation and planning. In real-world applications, optimization problems are generally hard, non-linear, discrete and/or large scale. State-of-the-art methods rely on human expertise to design specialized heuristics for different generic classes of problems and settings. They cannot leverage the specifics of the distribution of problem instances as they are repeatedly encountered in practice in a given context, although that information is usually easy to collect and made available for processing.
We are interested in data-driven optimization, and in particular learning underlying patterns in the problem data to be exploited in the optimization process. Our research topics are at the intersection of machine learning and (stochastic) optimization, spanning end-to-end learning, heuristic learning and hybrid methods that combine learning components and classical approaches. While solutions produced by ad-hoc solvers may be used as supervision signals, this often comes at a cost. We are interested in reinforcement learning techniques as a powerful framework for learning without relying on the availability of solved instances.
- On the generalization of neural combinatorial optimization heuristics, Sahil Manchanda, Sofia Michel, Darko Drakulic, Jean-Marc Andreoli, ECML-PKDD Sept 2022
- Routing in multimodal transportation networks with non-scheduled lines,Darko Drakulic, Christelle Loiodice, Vassilissa Lehoux, International Symposium on Experimental Algorithms (SEA), July 2022.
- Itineraries evaluated and ranked using fuzzy logic, Lizhi Wang, Vassilissa Lehoux, Marie-Laure Espinouse, Van-Dat Cung, Journal of Intelligent Transportation Systems, Volume 26, 2022 – Issue 4
- Structured Time Series Prediction without Structural Prior, Darko Drakulic, Jean-Marc Andreoli, arXiv 2022
- Simple and Effective Balance of Contrastive Losses, Arnadu Sors, Rafael Sampaio de Rezende, Sarah. Ibrahimi, Jean-Marc Andreoli, arXiv 2021
- Simple and effective balancing of contrastive losses, Arnaud Sors, Rafael Sampaio De Rezende, Sarah Ibrahimi, Jean-Marc Andreoli, arXiv 2021
- Transfer customization with the trip-based public transit routing algorithm, Vassilissa Lehoux, Christelle Loiodice, Symposium on Algorithmic Approaches for Transportation Modelling, Optimization, and Systems, September, 2021
- The use of fuzzy logic in various combinatorial optimization problems, Darko Drakulic, Aleksandar Takaci, Miroslav Marić, Artificial Intelligence: Theory and Applications (pp.137-153), July 2021
- Exploring the Decision component of the Activation-Decision-Construction-Action Theory for gain and loss facing scenarios, Tei Laine, Tomi Silander, MathPsych/ICCM, July 2021