作者: D. Eppstein , M. S. Paterson , F. F. Yao
DOI: 10.1007/PL00009293
关键词: Ball tree 、 Random graph 、 Null graph 、 Combinatorics 、 Mathematics 、 Random geometric graph 、 Random regular graph 、 Fixed-radius near neighbors 、 Nearest neighbor graph 、 Line graph 、 Discrete mathematics
摘要: The "nearest-neighbor" relation, or more generally the "k-nearest-neighbors" defined for a set of points in metric space, has found many uses computational geometry and clustering analysis, yet surprisingly little is known about some its basic properties. In this paper we consider natural questions that are motivated by geometric embedding problems. We derive bounds on relationship between size depth components nearest-neighbor graph prove probabilistic properties k-nearest-neighbors random points.