Skip to Main content Skip to Navigation

Propagation et réduction des incertitudes dans les modèles de changement d’occupation des sols

Abstract : This thesis presents an approach for propagating and reducing uncertainty related to the process of land cover change (LCC). LCC models have, almost, two sources of uncertainty which are the uncertainty related to model parameters and the uncertainty related to model structure. Furthermore, uncertainties can be subdivided into two types: aleatory and epistemic. The former comes from the natural variability, while the latter represents a lack of knowledge. The main problem is that classical probability theory does not make a clear distinction between aleatory and epistemic uncertainties in the way they are represented, i.e., both of them are described with a probability distribution. The aim of this thesis is to propagate and reduce the uncertainty related to LCC models using belief function theory. More precisely, the proposed approach is divided into six main steps: 1) a step for parameter uncertainty identification used to identify the uncertain input parameters, their types of uncertainty and their correlations, 2) a step for parameter modeling used to model and to transform all uncertain parameters in belief functions framework, 3) a step for parameter uncertainty propagation used to propagate aleatory and epistemic uncertainty of input parameters with considering correlations, 4) a step for model structure modeling and propagating used to model and propagate the uncertainty associated to model structure, 5) a step for sensitivity analysis performed to test the effect of uncertainty sources, and 6) a step for estimation based on confidence limits of KolmogorovSmirnov used to optimize the selected ones. In this thesis, the proposed approach is applied to an urbanization scenario. The present and future changes of a case study were studied using multi-temporal Landsat satellite images. These data are used for the preparation of a prediction map of the year 2025. Results show that the proposed approach based on the belief function theory has a potential for reducing uncertainty in LCC modeling.
Complete list of metadatas

Cited literature [151 references]  Display  Hide  Download
Contributor : Ahlem Ferchichi <>
Submitted on : Tuesday, January 2, 2018 - 4:25:01 PM
Last modification on : Friday, October 23, 2020 - 4:39:18 PM
Long-term archiving on: : Tuesday, April 3, 2018 - 12:20:34 PM


Files produced by the author(s)


  • HAL Id : tel-01667936, version 1



Ahlem Ferchichi. Propagation et réduction des incertitudes dans les modèles de changement d’occupation des sols . Intelligence artificielle [cs.AI]. Université de la Manouba 2017. Français. ⟨tel-01667936⟩



Record views


Files downloads