上财保险精算班生存模型课件.pdf

0 INTRODUCTION 1 Chapter 7 Survival models 0 Introduction • Deal with the random lifetimes by treat the future lifetime of an individual as a contin- uous random variable F (t) = P [T ≤ t] S (t) = P [T t] = 1 − F (t) • Derive useful results that are the building blocks of actuarial work relating to human mortality 0 INTRODUCTION 2 • Can extend mortality to other problems, e.g. the onset of illness • Chapter 8 (Estimating the lifetime distribu- tion function) • Chapter 9 (Cox regression model) • Chapter 10 (Binomial and Poisson models) • Chapter 11 (Exposed to risk) • Chapter 12 (Graduation and statistical tests) • Chapter 13 (Methods of graduation) 0 INTRODUCTION 3 • The contents of this chapter are 1 A simple model of survival 2 Expected future lifetime 3 Some important formulae 4 Simple parametric survival models 5 The Gompertz and Makeham laws of mortality 1 A SIMPLE MODEL OF SURVIVAL 4 1 A simple model of survival 1.1 Future lifetime • Assumption The future lifetime of a new-born baby is a random variable, denote T , which is con- tinuously distributed on an interval [0, ω] where 0 ω ∞. • For practical work, typical values of ω are in the range 100 − 120. 1 A SIMPLE MODEL OF SURVIVAL 5 • Definition F (t) = P [T ≤ t] is the distribution func- tion of T . S (t) = P [T t] = 1 − F (t) is the surviva