Apprentissage : Modèles et Algorithmes

The main area of research of our team are the models and methods allowing a system to adapt to its environment and provide pertinent responses based on available information. This domain covers machine learning (algorithms able to acquire knowledge based on empirical data) and the modelization of human learning (both in teaching situations and through social interactions). Our team has contributions in statistical learning theory, in algorithmic developments for classification of numerical and symbolic data and in cognitive modelling. Our approaches combine different and complementary representation paradigms, which are seldom found in a same group, like Inductive Logic Programming, neural networks, Latent Semantic Analysis, reinforcement learning, etc.