基于点过程模拟的时空级联模式统计挖掘方法
发布时间：2018-07-17 04:00
【摘要】：从时空统计的角度,将时空级联模式的频繁度评价建模为多元独立分布零假设下的显著性判别问题,提出一种基于点过程模拟的时空级联模式统计挖掘方法。首先,采用时空点过程模拟每类地理要素的观测数据集,构建显著性判别的零模型;其次,通过蒙特卡洛模拟获取零假设下每种候选时空级联模式频繁度的实验分布;最后,对候选模式的观测频繁度进行显著性检验,识别显著的时空级联模式。研究结果表明:本文方法能够用于有效识别地理要素间的时空级联模式,且避免了挖掘结果对频繁度阈值设置的依赖。
[Abstract]:From the point of view of space-time statistics, the frequency evaluation of spatio-temporal cascaded patterns is modeled as a significant discrimination problem under the assumption of zero multivariate independent distribution, and a statistical mining method of spatio-temporal cascaded patterns based on point process simulation is proposed. Firstly, the spatio-temporal point process is used to simulate the observational data set of each kind of geographical elements, and the zero model of significant discrimination is constructed. Secondly, the experimental distribution of the frequency of each candidate spatio-temporal cascade model under zero assumption is obtained by Monte Carlo simulation. The observational frequency of candidate models was tested to identify significant spatio-temporal cascade patterns. The results show that the proposed method can be used to identify spatio-temporal cascade patterns between geographical elements effectively and avoid the dependence of mining results on frequency threshold setting.
【作者单位】： 中南大学地球科学与信息物理学院;
【基金】：国家自然科学基金资助项目(41471385) 湖南省自然科学杰出青年基金资助项目(14JJ1007)~~
【分类号】：P208
本文编号：2128919
[Abstract]:From the point of view of space-time statistics, the frequency evaluation of spatio-temporal cascaded patterns is modeled as a significant discrimination problem under the assumption of zero multivariate independent distribution, and a statistical mining method of spatio-temporal cascaded patterns based on point process simulation is proposed. Firstly, the spatio-temporal point process is used to simulate the observational data set of each kind of geographical elements, and the zero model of significant discrimination is constructed. Secondly, the experimental distribution of the frequency of each candidate spatio-temporal cascade model under zero assumption is obtained by Monte Carlo simulation. The observational frequency of candidate models was tested to identify significant spatio-temporal cascade patterns. The results show that the proposed method can be used to identify spatio-temporal cascade patterns between geographical elements effectively and avoid the dependence of mining results on frequency threshold setting.
【作者单位】： 中南大学地球科学与信息物理学院;
【基金】：国家自然科学基金资助项目(41471385) 湖南省自然科学杰出青年基金资助项目(14JJ1007)~~
【分类号】：P208
【相似文献】
相关期刊论文 前1条
1 裴韬;李婷;周成虎;;时空点过程:一种新的地学数据模型、分析方法和观察视角[J];地球信息科学学报;2013年06期
，本文编号：2128919
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2128919.html
