预测死亡率分布的一个最小叉熵模型

A minimum cross entropy model for forecasting mortality

被保人的死亡率分布是确定寿险费率的一个重要依据,而根据其生活的环境、时间预测被保人的死亡率是保险精算研究中的一个热点问题.基于最小叉熵原理,建立了预测被保人死亡率分布的一个模型——最小叉熵模型,该模型以叉熵函数作为目标函数,以被保人的预期寿命作为约束条件,通过最小化叉熵预测被保人的死亡率.以从事特殊职业的被保人为研究对象,通过最小叉熵模型计算了该类人的死亡率.该方法计算简便,具有较好的客观性和实用性,为死亡率预测研究提供了一种有效的新方法.

Mortality distribution is an important criterion to determine the ratemaking of life insurance, and it has been attracting the significant interest of researchers to forecast the mortality of people living in different environments or periods in recent years. Based on the minimum cross entropy theory, a new model, which can forecast mortality distribution by minimizing the cross entropy function subject to the future life expectancy of insurants, is proposed. In addition, the mortality of the people working in the particular environment is calculated. From the process it can be seen that the model can be easily solved by simple calculation, and it is efficient.