A Fuzzy Approach for Modeling Human Mortality and Fertility
Tarih ve Saat
3 Ekim 2016 - 14:00
Abstract: Modeling and forecasting human mortality and fertility are significant research topics in several disciplines because these demographic rates are fundamental in financial planning and social policy decisions. Among various techniques, Lee Carter (LC) model is one of the most popular stochastic method in human mortality and fertility modeling. The original LC method for mortality modeling was fuzzified to eliminate the assumptions related with homoscedasticity and the ambiguity on the size of the error term variances by Koissi and Shapiro. The fuzzified LC model make use of ordinary least squares (OLS) technique, which prevents the model to capture the possible fluctuations in data. To overcome this issue, in this study, a modified version of fuzzy LC model utilizing singular value decomposition (SVD) technique for modeling human mortality and fertility is proposed. The proposed method is composed of two phases: Phase I – the fuzzification of observed crisp demographic rates, and Phase II – estimation phase. Phase I makes use of fuzzy regression technique based on minimum fuzziness criterion and linear programming, while Phase II includes solving an unconstrained nonlinear optimization model. For illustration purposes, the proposed method is applied to mortality and fertility data of Finland. Numerical outputs show that proposed method gives statistically better fits for both demographic rates compared to existing method.