带有约束条件的函数型数据的拟合——以北京市城镇最低生活保障为例

Fitting With the Constraints of Functional Data——Taking as an Example of Urban Minimum Living Allowance for Beijing

城镇低保问题一直以来被社会广泛关注,精确的拟合分布函数不仅对计算低保水平及覆盖范围起到重要的作用,同时有利于完善收入分配体制。我们尝试一种基于样条拟合和带有约束条件的函数型数据理论相结合的方法,首先将观测的数据转换为函数型数据,通过B-样条逼近数据的离散值,然后利用带有约束条件的函数型数据的分析方法及表现形式对拟合的函数加以约束和限制,不仅能满足分布函数的性质,而且能很好地兼顾拟合优化和曲线的光滑度。

Society has been widespread concern urban minimum living security issues, Precise fitting distribution function is calculated only for subsistence level and coverage plays an important role. At the same time help improve the income distribution system. We try a spline fit and function with constraints Data based on a combination of theory. First, the data is converted to a functional data of the observed. Approximation of discrete values of data through B - spline. Then use analytical methods and forms with the constraints of functional data to be constraints and restrictions on the fitting function. Not only to meet the nature of the distribution function, and can be a good fit both optimization and curve smoothness.