Approximation to mortality curves with a non-parametric proposal: The case of the Heligman and Pollard model

A non-parametric method is proposed to approximate mortality curves with emphasis on those coming from the Heligman and Pollard model (HP), where loss functions and controlled smoothness are applied such that the trend of the observed mortality data is as close as possible to the curve of mortality...

Descripción completa

Detalles Bibliográficos
Autores: Silva, Eliud, Ovin, Adrián
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2018
País:México
Institución:EL COLEGIO DE MÉXICO
Repositorio:Estudios Demográficos y Urbanos
Idioma:español
OAI Identifier:oai:oai.estudiosdemograficosyurbanos.colmex.mx:article/1805
Acceso en línea:https://estudiosdemograficosyurbanos.colmex.mx/index.php/edu/article/view/1805
Access Level:acceso abierto
Palabra clave:Heligman and Pollard model
smoothness
smoothing parameter
trends
mortality curve
generalized least squares
modelo de Heligman y Pollard
suavidad
parámetro de suavizamiento
tendencias
curva de mortalidad
mínimos cuadrados generalizados
Descripción
Sumario:A non-parametric method is proposed to approximate mortality curves with emphasis on those coming from the Heligman and Pollard model (HP), where loss functions and controlled smoothness are applied such that the trend of the observed mortality data is as close as possible to the curve of mortality produced by the HP model. It can be appreciated with several estimates that the best fit results are obtained for male mortality, and even more so after 12 years old. Using this perspective, it is provided a reference range of the smoothing parameters to approximate those mortality curves that would be obtained after applying the HP model.