Estimation of Longevity using Survival Functions

Abstract

Estimation of life expectancy and longevity are frequently used in several fields basically on convectional model. The study considers Longevity estimation models and variances as alternative to conventional model incorporating non-parametric, semi-parametric and parametric with relevant survival functions to formulate longevity models that will accommodate expected variables in the estimation of life expectancy. These models can be used to analyze both grouped and individual data. We consider Kaplan Meier model as nonparametric methods, Cox proportional model as semi parametric methods. The parametric methods considered are exponential, weibull and gompertz proportional models which are commonly used for analyzing discrete data. The derived variances of KM and Cox proportional longevity estimator models are equivalent this confirm the result of major work done in this field. The other derived longevity models have difference variances due to their peculiarities. We shall consider these models empirically in another paper.