Mortality Modelling and Forecasting using Cross-Validation Techniques
DOI:
https://doi.org/10.26481/marble.2015.v1.92Abstract
This paper investigates how accurately several predictive models perform in the field of mortallity modelling and forecasting, using cross-validation techniques. The main models used in this paper are the well-known Lee-Carter model and the Heligman-Pollard model. Furthermore, some extensions will be investigated: a distinction between the male and female population, extension to a few other countries, the effects of increasing the size of the ''training set'' and forecasting without re-estimating k_t for the Lee-Carter model and a weighted estimation for the Heligman-Pollard model.References
Booth, H., R.J. Hyndman, et al. (2006). Lee-Carter mortality forecasting: a multi-country comparison of variants and extensions". In: Demographic Research 15.9, pp. 289-310. doi: 10.4054/DemRes.2006.15.9.
Booth, H. and L. Tickle (2008). Mortality Modelling and Forecasting: a Review of Methods". In: Annals of Actuarial Science 3, pp. 3-43. doi: 10.1017/S1748499500000440.
Heligman, L. and J.H. Pollard (1980). The age pattern of mortality". In: Journal of the Institute of Actuaries 107.434, pp. 49-80.
Hyndman, R.J. (2006). demography: Forecasting mortality and fertility data, R package. url: http://www.robhyndman.info/Rlibrary/demography.
Lee, R.D. and L.R. Carter (1992). Modelling and forecasting U.S. mortality". In: Journal of the American Statistical Association 87.419, pp. 659-671.
Refaeilzadeh, P., L. Tang, and H. Liu (2009). Cross Validation". In: Encyclopedia of Database Systems. Springer.
Wilmoth, J.R. (1993). Computational methods for tting and extrapolating the Lee-Carter model of mortality change. Technical report, Department of Demography, University of California, Berkeley.
