适应性网络空间抽样的设计及估计研究

Design and Estimation of Adaptive Web Space Sampling

随着大数据背景下抽样环境日趋复杂化,特别是3S技术(遥感技术、地理信息系统和全球定位系统)日趋成熟,越来越多的社会经济问题涉及空间抽样,其样本呈现出规模相对稀少、分布不均匀、局部聚集的特征,使得传统抽样调查面临着严重挑战。本文介绍了适应性抽样技术应用于空间网络环境的基本原理、主要操作步骤和马尔可夫链蒙特卡罗估计推断,并以广州市天河区的商户抽样为例讨论实际操作中应注意的问题,以期为流动人口、环境污染、区域经济研究等方面的调查提供理论支撑和实证方法参考。

With the complexity of the sampling environment in the background of big data,especially the maturing of 3 S technology(remote sensing technology,geographic information system and global positioning system),more and more socio-economic problems involve spatial sampling,which is characterized by relatively sparse size,uneven distribution and local aggregation,so that the traditional sampling survey faces serious challenges.Our paper introduces an innovation adaptive sampling technology in the space network environment,such as the basic principle,operational procedures,MCMC estimation based on the example of Tianhe District in Guangzhou City.Our findings provide helpful theoretical support and empirical reference for the space sampling.