Skip to Main content Skip to Navigation
Theses

Méthodes statistiques pour l'évaluation du risque alimentaire

Abstract : Contaminants and natural toxicants such as mycotoxins may be present in several food items, which may be considered as dangerous for human health if the cumulative intake remains above the toxicological safe references. We focus on the estimation of the risk, defined as the probability for exposure to exceed a tolerable intake on a long term basis. Extreme value theory allows to quantify very low risk. In others cases, the empirical estimator of the risk is written as a generalised U-statistic, which allows to derive its asymptotic properties. Statistical developments are used to model the left censorship of the analytical data. Finally, an econometric model aiming at decomposing household quantities into individual quantities is used to propose a new method for the quantification of the long term risk integrating the possible accumulation and slow degradation of the contaminant in the human organism.
Document type :
Theses
Complete list of metadatas

Cited literature [9 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00139909
Contributor : Jessica Tressou <>
Submitted on : Wednesday, April 4, 2007 - 8:06:16 AM
Last modification on : Friday, October 23, 2020 - 4:52:10 PM
Long-term archiving on: : Friday, November 25, 2016 - 3:29:54 PM

Identifiers

  • HAL Id : tel-00139909, version 1
  • PRODINRA : 148894

Collections

Citation

Jessica Tressou. Méthodes statistiques pour l'évaluation du risque alimentaire. Mathématiques [math]. Université de Nanterre - Paris X, 2005. Français. ⟨tel-00139909⟩

Share

Metrics

Record views

7732

Files downloads

12613