View Submission - CMStatistics

B0859
**Title: **Why does the human die: Cohort-wise mortality prediction under the survival energy hypothesis
**Authors: **Yasutaka Shimizu - Waseda University (Japan) **[presenting]**

**Abstract: **A quite new methodology is proposed for the mortality prediction. We assume there exists the survival energy in human beings: we are born with a certain initial quantity of survival energy, and the energy changes stochastically in time. The human dies if the energy decreases and hits the level zero, that is, the time of death is the first hitting time of the survival energy processes. We suppose the survival energy follows a diffusion process and define the mortality as the first hitting time distribution of the process. We estimate the unknown parameters in the stochastic differential equations via the least squares estimation by fitting the model to the empirical hitting time distribution function using the historical data from ``Human Mortality Database''. We shall illustrate by the real data analysis that such a ``structural approach'' to the human energy gives a surprisingly good prediction for long-future's mortality.