Two novel methods of life expectancy estimation, applied to various annual reported demographic datasets, are proposed. First, for datasets that fully recorded birth date and death date of all dead individuals, we rely on the well-known Kaplan–Meier estimation method to provide an accurate estimation framework of life expectancy. Our proposed method can be used as a gold standard in the accuracy investigation of other life expectancy estimation methods. The method can be applied for small areas, where complete mortality data are regularly produced by routine annual surveys. The second new created method, called as local parametric method, based on the theoretical background of survival process with local parametric Weibull distributions, estimates life expectancy using abridged survival data. Experiments on real longitudinal datasets show the new method provides very exact life expectancy estimations for 10 among 15 one-year datasets, whilst the method of Chiang often yields overestimations.

提出了两种新的预期寿命估计方法，适用于各种年度报告的人口数据集。首先，对于完全记录所有死亡个体的出生日期和死亡日期的数据集，我们依靠著名的 Kaplan-Meier 估计方法来提供准确的预期寿命估计框架。我们提出的方法可以用作其他预期寿命估计方法的准确性调查的黄金标准。该方法可用于小区域，通过例行年度调查定期生成完整的死亡率数据。第二种新创建的方法称为局部参数方法，它基于具有局部参数 Weibull 分布的生存过程的理论背景，使用删节的生存数据估计预期寿命。