Skip to Main content Skip to Navigation
Journal articles

Le boosting : une méthode de classification non paramétrique

Abstract : Boosting is a classification method used to assign an object to a predetermined class, according to the values obtained for several variables. It is a very general method, whieh can be used fbr a avide range of applications (economics, agronomy, sociology, geography, pedology, epidemiology, remote sensing, etc), particularly those in which the object is geographical space. The method is vert' powerful, and can discriminate between groups where other techniques such as linear methods would fail. However, it calls for considerable caution and a watchful eye on certain criteria such as the weight of observations. Lastly, it is easily integrated into geographical information systems.
Document type :
Journal articles
Complete list of metadata
Contributor : Migration Irstea Publications <>
Submitted on : Thursday, May 14, 2020 - 9:08:23 PM
Last modification on : Friday, April 16, 2021 - 3:29:23 AM


  • HAL Id : hal-02581438, version 1
  • IRSTEA : PUB00011626



M. Arnaud, Jean-Stéphane Bailly, G. Bourgeon, C. Puech. Le boosting : une méthode de classification non paramétrique. Revue Internationale de Géomatique, Lavoisier, 2002, 12 (4), pp.399-420. ⟨hal-02581438⟩



Record views