logo BDSP

Base documentaire


  1. Eliminating a sparse data problem in nonparametric bivariate survival analysis with an application to women's age at first marriage and age at first birth.

    Article - En anglais

    A new solution is proposed for a sparse data problem arising in nonparametric estimation of a bivariate survival function.

    Prior information, if available, can be used to obtain initial values for the EM algorithm.

    Initial values will completely determine estimates of portions of the distribution which are not identifiable from the data, while having a minimal effect on estimates of portions of the distribution for which the data provide sufficient information.

    Methods are applied to the distribution of women's age at first marriage and age at birth of first child, using data from the Current Population Surveys of 1975 and 1986.

    Mots-clés Pascal : Femme, Homme, Démographie, Estimation non paramétrique, Algorithme EM, Condition initiale, Etude théorique, Analyse mathématique, Age

    Mots-clés Pascal anglais : Woman, Human, Demography, Non parametric estimation, EM algorithm, Initial condition, Theoretical study, Mathematical analysis, Age

    Logo du centre Notice produite par :
    Inist-CNRS - Institut de l'Information Scientifique et Technique

    Cote : 93-0137503

    Code Inist : 002B30A01A1. Création : 199406.