Skip to Main content Skip to Navigation
Theses

Surveillance dynamique de mesures

Abstract : The measures are daily actions, they give us a lot of information and allow us to make decisions. The analysis of measures can allow us to learn more about our environment, but the error of a measure can have important consequences in certain areas. In a first part, we propose, thanks to the study of blood test measurements carried out at the CHU of Clermont-Ferrand, a procedure for detecting deviations from medical biology laboratory analyzers based on patient analysis measurements. After a descriptive analysis of the data, the method put in place, using methods of detection of breaks of time series, is tested for simulations of breaks representing offsets, imprecision or drifts of machine for different measured biological parameters. The method is adapted for two scenarios: when the patient's hospital service is known or not. The study is supplemented by an analysis of the impact of measurement uncertainty on patient analyses. In a second part we study measurements of volcanic ash forms made at “Laboratoire Magmas et Volcans” of the Clermont Auvergne University, in order to determine a link between the collection locations and the forms of the particles. After showing the dependence between these parameters, we propose, using a classification method, a grouping of particles representing different populations depending on the distance between the collection locations and the volcano crater.
Document type :
Theses
Complete list of metadatas

Cited literature [73 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02866517
Contributor : Abes Star :  Contact
Submitted on : Friday, June 12, 2020 - 3:07:11 PM
Last modification on : Saturday, June 13, 2020 - 3:41:42 AM

File

2019CLFAC091_MIALLARET.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02866517, version 1

Citation

Sophie Miallaret. Surveillance dynamique de mesures. Statistiques [math.ST]. Université Clermont Auvergne, 2019. Français. ⟨NNT : 2019CLFAC091⟩. ⟨tel-02866517⟩

Share

Metrics

Record views

138

Files downloads

125