基于自相关的我国农民人均纯收入的空间关联规则挖掘

Autocorrelation Based Spatial Incident Rule Mining of Chinese Farmers' Per Capita Net Income

空间自相关是空间统计的一种分析方法,能在设定显著性水平下研究相邻区域位置属性或现象之间的相关性,是分析区域变量的空间结构特征的有效手段。本文通过时2000～2004年我国农民人均纯收入进行全局和局部空间自相关分析,将区域分布划分为高收入地区、潜在高收入地区及低收入地区,揭示了全国范围内各省农民人均纯收入在整体上呈现出显著的空间正相关特性,其空间聚类趋势明显。

The spatial autocorrelation is an analytical method of spatial statistics. It can study the correlativity of the location attribute and phenomenon between the neighboring regions under a given significance level, and is an efficient means for analyzing the spatial structure features of the area variable. Using the spatial autocorrelation technique, this paper conducts the overall and local analysis of the Chinese farmers＇ per capita net income from 2000 - 2004, and divides the study areas into three categories of high income area, potential high income area and low income area. This study shows that in the whole country, the spatial correlation of the farmer＇s per capita net income between different provinces is obviously positive, and there is a clear tendency of spatial clustering.