Research Themes

Project-Team TAO Thème Apprentissage & Optimisation common team to INRIA Saclay and LRI

The cross-fertilisation of machine learning, knowledge discovery and evolutionary computation is the main motivation for the PCRI TAO project, created in 2003 and merging most of the I&A group with part of the former INRIA Fractales group. On the one hand, evolutionary computation offers unrivalled opportunities for machine learning and knowledge discovery, specifically for investigating non-convex learning criteria and controlling the trade-off between the computational resources needed and the hypothesis quality (any-time algorithms). On the other hand, evolutionary computation will highly benefit from a principled, machine learning-based, exploitation of the wealth of data generated by an EC application. This data is at the moment poorly exploited, if at all.

The I&A - TAO group involves six highly coupled research themes that are detailed in their respective pages (see below).

• Autonomic Computing
• Crossing the Chasm
• Complex Systems
• Continuous Optimization
• Optimal Decision Making under Uncertainty
• Reservoir Computing

• Read the 4-years report written for the INRIA 2003-2007 external evaluation.
• Browse the yearly activity reports on the INRIA web site
• Directly download the corresponding PDF files
  • 2007
  • 2006
  • 2005
  • 2004
• Miscellaneous
  • Transparents de la présentation de TAO au Comdir INRIA du 6 juin 2006

• A very short introduction to Evolutionary Computation (A.E. Eiben and M. Schoenauer, Evolutionary Computation, Information Processing Letter 82(1):1-6, 2002.
• Une introduction un peu plus complète, en français