Team tao

Overall Objectives
Scientific Foundations
Application Domains
New Results
Contracts and Grants with Industry
Other Grants and Activities

Section: New Results

Crossing the Chasm

Participants : Alejandro Arbelaez, Anne Auger, Djalel Benbouzid, Robert Busa-Fekete, Alvaro Fialho, Nikolaus Hansen, Balázs Kégl, Marc Schoenauer, Michèle Sebag.

Adaptive Operator Selection (AOS):

within the Microsoft-INRIA joint lab, Alvaro Fialho completed his PhD [2] . New algorithms based on Multi-Armed Bandits (MAB) have been proposed: The Sliding MAB has been introduced, and thoroughly compared with previous MAB-based AOS [15] ; the best-performing AOS to-date removes the dependency on fitness values and leans toward comparison-based AOS [67] using as a reward the Area Under the Curve, that is popular in Machine Learning; the paradigm has also been validated in a different context, namely that of Differential Evolution for Continuous Optimization [66] , [65] . A book chapter, written with our colleagues from Angers, surveys the AOS domain (though written after our latest propositions) [111] .

Adaptive Constraint Programming:

Previous work on Automatic Heuristic Selection [126] has been applied to the problem of Protein Structure Prediction [22] . Furthermore, a new procedure to take advantage of idle time between optimization to learn from previously tackled instances has been proposed [24] .

Improved AdaBoost:

We showed how to accelerate AdaBoost using MABs [54] and applied the technique for ranking [55] . This algorithm was ranked 5th in the 2010 Yahoo Ranking Challenge.

Adaptive Encoding:

All work pertaining to Continuous Optimization (see following Section 6.4 ) and dealing with CMA-ES variants also belongs to this theme, from the derivation of new algorithms [52] , [25] , [76] , [75] , [74] to the development of new applications [119] .


Logo Inria