Carlo Giovanni Camarda imparteix la conferencia “Smooth Constrained Mortality Forecasting” en el marc dels Colloquium CED
Organiza: Centre d’Estudis Demogràfics
Lugar: Sala Àngels Torrents, CED
Hora: 11:00 - 12:30
Previ al Col·loqui, de 10:40 a 11:00h, esteu tots convidats a l’Espai del cafè.
Carlo Giovanni Camarda (Investigador i Cap de la Unitat de Recerca «Mortality, Health and Epidemiology», Institut National d’Etudes Démographiques-INED, França).- Smooth Constrained Mortality Forecasting.
Abstract.- Smooth Constrained Mortality Forecasting.
Mortality can be forecast by means of parametric models, principal component methods, and smoothing approaches. These methods either impose rigid modeling structures or produce implausible outcomes. During the talk, we first introduce in a gently, albeit rigorous manner demographic forecasting models as well as a well established smoothing approach for mortality data: P-splines.
Then we propose a novel approach for forecasting mortality that combines P-splines and prior demographic information. We constrain future smooth mortality patterns to lie within a range of valid age profiles and time trends, both computed from observed patterns. Within a P-spline framework, we enforce shape constraints through an asymmetric penalty approach on forecast mortality. Hence we call the proposed approach a CP-spline model.
Moreover, we properly integrate infant mortality in a smoothing framework so that the mortality forecast covers the whole age range. The proposed model outperforms the plain smoothing approach as well as commonly used methodologies while retaining all the desirable properties that demographers expect from a forecasting method, e.g., smooth and plausible age profiles and time trends. We illustrate the proposed approach to mortality data for Danish females and US males.
The proposed methodology offers a new paradigm in forecasting mortality, and it is an ideal balance between pure statistical methodology and traditional demographic models. Prior knowledge about mortality development can be conveniently included in the approach, leading to large flexibility. The combination of powerful statistical methodology and prior demographic information makes the proposed model suitable for forecasting mortality in most demographic scenarios.