Skip to Main content Skip to Navigation
Theses

Méthodes d'Extraction de Connaissances à partir de Données (ECD) appliquées aux Systèmes d'Information Géographiques (SIG)

Abstract : During this PhD thesis, we have studied methods for Knowledge Discovery in Databases (KDD) applied to Geographic Information Systems (GIS). We have improved both classical KDD methods (Data Clustering, Cluster Visualization) and spatial KDD methods linked with spatial analysis methods (Spatial Smoothing, Hot Spot Extraction, Spatial Partitionning). We have worked in GÉOBS, a company expert in spatial data analysis. So our KDD methods have been implemented and tested with data sets provided by GÉOBS in relation with Economic Development, Geomarketing, Risk Analysis, Environnement, Health, etc. This report gives a wide point of view on a range of analysis methods and their related problems. It points up the complementarity between theses methods which can be connected either in a technical way or in a user way. Eventually, this work was very enriching because it has concerned many problems and as many KDD tools.
Document type :
Theses
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00101491
Contributor : Christophe Candillier <>
Submitted on : Wednesday, September 27, 2006 - 1:17:25 PM
Last modification on : Friday, October 23, 2020 - 4:33:26 PM
Long-term archiving on: : Tuesday, April 6, 2010 - 1:14:48 AM

Identifiers

  • HAL Id : tel-00101491, version 1

Collections

Citation

Christophe Candillier. Méthodes d'Extraction de Connaissances à partir de Données (ECD) appliquées aux Systèmes d'Information Géographiques (SIG). Interface homme-machine [cs.HC]. Université de Nantes, 2006. Français. ⟨tel-00101491⟩

Share

Metrics

Record views

1086

Files downloads

5406