摘要
文献 [1]、[2 ]从理论和实践上证明了信息扩散估计是一种很好的估计。但寻找更合适的窗宽 ,是扩散估计的最大难题之一 ,也是影响扩散估计精度的最关键因素之一。在深入讨论信息扩散估计的基本大样本性质的基础上 ,提出了最优窗宽的理论和算法 ,较圆满地解决了窗宽问题。
Information spread estimation (ISE) not only has fine robustness, but also it is very simple and convenient for application. However, the estimation of the density function depends on the window-width. If the best estimation result is wanted to obtain, the optimum window-width must be choosen. How to acquire the optimum window-width is undoubtedly one of the most difficult problem in information spread estimation, and one of the most important factors influencing the precision in the estimation. Based on a detailed study in the nature of big samples in normal information spread estimation,this paper presents a theory and algorithm on how to acquire the optimum window-width . Then the problem of how to choose the optimum window-width has been successfully solved.
出处
《测绘科学》
CSCD
2001年第1期16-19,共4页
Science of Surveying and Mapping
基金
国家自然科学基金资助项目! (4 98740 0 2 )
国家测绘局测绘科技发展基金资助项目! (990 10 )
