Evolutionary predictive modelling for flash floods

Wilfried Segretier 1 Martine Collard 2 Manuel Clergue 2
LAMIA - Laboratoire de Mathématiques Informatique et Applications
Abstract : Modelling techniques for river hydrologic forecasting systems have taken advantage of machine learning methods especially for flood prediction. But current solutions, mostly based on artificial neural networks do not always meet end users requirements on the readability and the understandability of predictive models. In this paper, we present a new version of our original solution based on the concept of aggregate variables in order to predict flash flood events from observed water level and/or rain measurements, particularly in the context of caribbean watersheds in which flash flood are much uncertain. We combine aggregate variables in juries. Juries of aggregate variables are trained and tested using a typical 10-fold cross validation scheme. Best juries are searched through an evolutionary approach that is optimized. Different parameters are set up like aggregation periods and jury sizes to prove the efficiency of the proposed approach compared to classical solutions.
Type de document :
Communication dans un congrès
Congress on Evolutionary Computation (CEC) 2013, Jun 2013, Cancun, Mexico. pp.844 - 851, 2013, <10.1109/CEC.2013.6557656>
Liste complète des métadonnées

Contributeur : Wilfried Segretier <>
Soumis le : mardi 23 juillet 2013 - 21:52:44
Dernière modification le : jeudi 14 avril 2016 - 01:05:55




Wilfried Segretier, Martine Collard, Manuel Clergue. Evolutionary predictive modelling for flash floods. Congress on Evolutionary Computation (CEC) 2013, Jun 2013, Cancun, Mexico. pp.844 - 851, 2013, <10.1109/CEC.2013.6557656>. <hal-00847582>



Consultations de la notice