Robust forecasting of mortality and fertility rates: A functional data approach

A new method is proposed for forecasting age-specific mortality and fertility rates observed over time. This approach allows for smooth functions of age, is robust for outlying years due to wars and epidemics, and provides a modelling framework that is easily adapted to allow for constraints and other information. Ideas from functional data analysis, nonparametric smoothing and robust statistics are combined to form a methodology that is widely applicable to any functional time series data observed discretely and possibly with error. The model is a generalization of the Lee-Carter (LC) model commonly used in mortality and fertility forecasting. The methodology is applied to French mortality data and Australian fertility data, and the forecasts obtained are shown to be superior to those from the LC method and several of its variants.

[1]  W. Bell,et al.  Using principal components in time series modeling and forecasting of age-specific mortality rates. , 1991 .

[2]  J HyndmanRob,et al.  Robust forecasting of mortality and fertility rates , 2007 .

[3]  Heng Tao Shen,et al.  Principal Component Analysis , 2009, Encyclopedia of Biometrics.

[4]  Ana M. Aguilera,et al.  Forecasting PC-ARIMA Models for Functional Data , 2002, COMPSTAT.

[5]  Elaine B. Martin,et al.  On principal component analysis in L 1 , 2002 .

[6]  Pin T. Ng,et al.  COBS: qualitatively constrained smoothing via linear programming , 1999, Comput. Stat..

[7]  R. Lee,et al.  Modeling and forecasting U.S. sex differentials in mortality. , 1992, International journal of forecasting.

[8]  I. Jolliffe.  Principal Component Analysis , 2002 .

[9]  D. G. Simpson,et al.  Robust principal component analysis for functional data , 2007 .

[10]  J. Simonoff.  Smoothing Methods in Statistics , 1998 .

[11]  E. Ziegel,et al.  Proceedings in Computational Statistics , 1998 .

[12]  Heather Booth,et al.  Evaluation of the Variants of the Lee-Carter Method of Forecasting Mortality: A Multi-Country Comparison , 2004 .

[13]  James O. Ramsay,et al.  Functional Data Analysis , 2005 .

[14]  Christophe Croux,et al.  A Fast Algorithm for Robust Principal Components Based on Projection Pursuit , 1996 .

[15]  Christophe Croux,et al.  High breakdown estimators for principal components: the projection-pursuit approach revisited , 2005 .

[16]  PIET DE JONG,et al.  Extending Lee–Carter Mortality Forecasting , 2006 .

[17]  Paul H. C. Eilers,et al.  Smoothing and forecasting mortality rates , 2004 .

[18]  R. Lee,et al.  Disaggregation in population forecasting: do we need it? And how to do it simply. , 1995, Mathematical population studies.

[19]  Gwilym M. Jenkins,et al.  Time series analysis, forecasting and control , 1972 .

[20]  Steven Haberman,et al.  Lee-Carter mortality forecasting: a parallel generalized linear modelling approach for England and Wales mortality projections , 2003 .

[21]  Vladimir M. Shkolnikov,et al.  Methods Protocol for the Human Mortality Database , 2002 .

[22]  Rob J Hyndman,et al.  Forecasting age‐specific breast cancer mortality using functional data models , 2007, Statistics in medicine.

[23]  B. Silverman,et al.  Smoothed functional principal components analysis by choice of norm , 1996 .

[24]  Leonie Tickle,et al.  Extending the Lee-Carter methodology of mortality projection , 2004 .

[25]  Frédéric Ferraty,et al.  Nonparametric models for functional data, with application in regression, time series prediction and curve discrimination , 2004 .

[26]  W. Bell,et al.  FORECASTING AGE SPECIFIC FERTILITY USING PRINCIPAL COMPONENTS , 1987 .

[27]  Andrew Harvey,et al.  Forecasting, Structural Time Series Models and the Kalman Filter , 1990 .

[28]  Wai-Sum Chan,et al.  Outlier analysis and mortality forecasting: The United Kingdom and Scandinavian countries , 2005 .

[29]  J. Ramsay,et al.  Some Tools for Functional Data Analysis , 1991 .

[30]  C. Croux,et al.  Generalizing univariate signed rank statistics for testing and estimating a multivariate location parameter , 1995 .

[31]  Ronald Lee,et al.  Modeling and forecasting U. S. mortality , 1992 .

[32]  R. Lee,et al.  Modeling and forecasting the time series of US fertility: age distribution, range, and ultimate level. , 1993, International journal of forecasting.

[33]  Rob J Hyndman,et al.  Prediction Intervals for Exponential Smoothing State Space Models , 2001 .

[34]  Rob J. Hyndman,et al.  Lee-Carter mortality forecasting: a multi-country comparison of variants and extensions , 2006 .

[35]  Rob J Hyndman,et al.  A state space framework for automatic forecasting using exponential smoothing methods , 2002 .

[36]  S. Wood.  Thin plate regression splines , 2003 .

[37]  J. Dauxois,et al.  Asymptotic theory for the principal component analysis of a vector random function: Some applications to statistical inference , 1982 .

[38]  Guoying Li,et al.  Projection-Pursuit Approach to Robust Dispersion Matrices and Principal Components: Primary Theory and Monte Carlo , 1985 .

[39]  Lon-Mu Liu,et al.  Joint Estimation of Model Parameters and Outlier Effects in Time Series , 1993 .

[40]  B. Ripley,et al.  Semiparametric Regression: Preface , 2003 .

[41]  M. Hubert,et al.  A fast method for robust principal components with applications to chemometrics , 2002 .

[42]  Richard A. Davis,et al.  Time Series: Theory and Methods , 2013 .

[43]  J. Maindonald,et al.  Applying Lee-Carter under conditions of variable mortality decline , 2002, Population studies.

[44]  Simon N. Wood,et al.  Monotonic Smoothing Splines Fitted by Cross Validation , 1994, SIAM J. Sci. Comput..