We propose a statistical method for estimating source apportionment of particulate matters based on source profiles with fluctuations.
Previous methods for estimating source apportionment are based on the assumption that source profiles are to be constant.
This assumption does not seem to describe precisely the real world, however.
For example, it is to be expected that the profiles of tobacco have sharp fluctuations, because they depend on various factors such as ventilation, kinds of cigarettes, and puffing.
Considering this situation, we develop a new model and a simple estimating method which allows source profiles to have fluctuations.
We illustrate our method using real data sets which are stratified with smoking status.
The proposed method provides reasonable results.
Mots-clés Pascal : Risque, Incidence, Evaluation, Epidémiologie, Mesure, Paramètre, Exposition, Homme, Prévalence, Analyse statistique, Maladie, Tabagisme, Tabagisme passif, Cardiopathie coronaire, Tumeur maligne, Appareil circulatoire pathologie, Appareil respiratoire pathologie, Bronchopulmonaire, Risque attribuable
Mots-clés Pascal anglais : Risk, Incidence, Evaluation, Epidemiology, Measurement, Parameter, Exposure, Human, Prevalence, Statistical analysis, Disease, Tobacco smoking, Passive smoking, Coronary heart disease, Malignant tumor, Cardiovascular disease, Respiratory disease, Bronchopulmonary
Notice produite par :
Inist-CNRS - Institut de l'Information Scientifique et Technique
Cote : 98-0007959
Code Inist : 002B30A01A1. Création : 17/04/1998.