Abstract | ||
---|---|---|
The conventional approach to estimate the global mean under preferential spatial sampling gets a larger deviation and further influences the precision of the subsequent model calculation and analysis. Zoning or declustering methods can effectively improve the estimation precision of preferential sampling. In this paper, we propose a novel method, which uses self-organizing dual-zoning method to estimate the global mean, in which the Self-Organizing Feature Map (SOFM) and the Voronoi diagram are utilized to realize classification and zoning. By comparing with arithmetic mean method, polygonal declustering method, and cell declustering method, we got that arithmetic mean method could not satisfy the special properties of the preferential sampling, and self-organizing dual-zoning method gets more accurate zoning results and more stable global means with different sample sizes and Feature Deviation Index (FDI). |
Year | DOI | Venue |
---|---|---|
2014 | 10.1080/10798587.2014.934591 | INTELLIGENT AUTOMATION AND SOFT COMPUTING |
Keywords | Field | DocType |
Preferential sampling, Global mean estimation, Self-organizing dual-zoning method, Polygonal declustering, Cell declustering | Zoning,Data mining,Polygon,Computer science,Arithmetic mean,Algorithm,Sampling (statistics),Artificial intelligence,Voronoi diagram,Sample size determination,Machine learning | Journal |
Volume | Issue | ISSN |
20 | 4 | 1079-8587 |
Citations | PageRank | References |
0 | 0.34 | 3 |
Authors | ||
6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xuhong Ren | 1 | 0 | 0.68 |
Ning Wei | 2 | 0 | 0.34 |
Bingbo Gao | 3 | 9 | 2.80 |
Yuchun Pan | 4 | 10 | 8.82 |
Qing Guo | 5 | 0 | 1.35 |
Yunbing Gao | 6 | 7 | 1.58 |