Skip to Main content Skip to Navigation
Conference papers

Imputation crédibiliste pour la prédiction de charge interne de joueurs de football

Rayane Elimam 1 Nicolas Sutton-Charani 2 Jacky Montmain 2 S. Perrey 1 
2 I3A - Informatique, Image, Intelligence Artificielle
LGI2P - Laboratoire de Génie Informatique et d'Ingénierie de Production
Abstract : The objective of this article is to compare three approaches of missing data management in a classification context with sequential labels at variable time steps. The data concern the longitudinal monitoring of twenty-seven professional soccer players in terms of training schedules and perceived internal load. One approach without imputation uses the age of the data, the other two approaches are based on uncertainty models at the missing data imputation step. The Evidential K-Nearest Neighbors (EKNN) is used for internal load prediction taking into account labels uncertainty, and the K-Nearest Neighbors for the approach without uncertain labels. Results show moderate prediction improvement for models based on imputed data uncertainty.
Document type :
Conference papers
Complete list of metadata
Contributor : Administrateur IMT - Mines Alès Connect in order to contact the contributor
Submitted on : Thursday, April 28, 2022 - 4:05:18 PM
Last modification on : Monday, May 2, 2022 - 9:08:58 AM


Files produced by the author(s)


  • HAL Id : hal-02969311, version 1


Rayane Elimam, Nicolas Sutton-Charani, Jacky Montmain, S. Perrey. Imputation crédibiliste pour la prédiction de charge interne de joueurs de football. LFA’2020 - 29èmes Rencontres Francophones sur la Logique Floue et ses Applications, Oct 2020, Sète, France. ⟨hal-02969311⟩



Record views


Files downloads