Imputation of partial non-responses in surveys with a self-organizing map based model

FESSANT ; MIDENET

Type de document
CHAPITRE D'OUVRAGE (CO)
Langue
anglais
Auteur
FESSANT ; MIDENET
Résumé / Abstract
An important issue with many data analysis problems is that different variables are sometimes missing. To address this issue we propose to use a neural network model based on kohonen's self-organizing maps. The learning process is performed on complete data. The learned weights are then used in the exploitation phase to fill in the missing values. Moreover the self-organizing map model is useful to represent the knowledge acquired while learning in a structured way that can be interpreted. The interest of the model is presented through an application of non-response treatment in surveys. Data come from the 1993-94 French national personal tranport survey.

puce  Accès à la notice sur le portail documentaire de l'IFSTTAR

  Liste complète des notices publiques de l'IFSTTAR