Universal consistency of the k-NN rule in metric spaces and Nagata dimension
Universal consistency of the k-NN rule in metric spaces and Nagata dimension
复制标题
度量空间和Nagata维中k-NN规则的普遍一致性
DOI:
10.1051/ps/2020018
复制
发表时间:
2020
期刊:
影响因子:
--
通讯作者:
Pestov Vladimir G.
中科院分区:
文献类型:
--
作者:
Collins Benoit;Kumari Sushma;Pestov Vladimir G.
Theknearest neighbour learning rule (under the uniform distance tie breaking) is universally consistent in every metric spaceXthat is sigma-finite dimensional in the sense of Nagata. This was pointed out by Cérou and Guyader (2006) as a consequence of the main result by those authors, combined with a theorem in real analysis sketched by D. Preiss (1971) (and elaborated in detail by Assouad and Quentin de Gromard (2006)). We show that it is possible to give a direct proof along the same lines as the original theorem of Charles J. Stone (1977) about the universal consistency of thek-NN classifier in the finite dimensional Euclidean space. The generalization is non-trivial because of the distance ties being more prevalent in the non-Euclidean setting, and on the way we investigate the relevant geometric properties of the metrics and the limitations of the Stone argument, by constructing various examples.
登录
查看更多内容
DOI:
10.3792/pja/1195523056
发表时间:
1963
期刊:
--
影响因子:
--
作者:
J. Nagata
通讯作者:
J. Nagata
DOI:
10.20381/ruor-3989
发表时间:
2015
期刊:
ArXiv
影响因子:
--
作者:
Stan Hatko
通讯作者:
Stan Hatko
影响因子:
1.2
作者:
P. Assouad;T. Q. D. Gromard
通讯作者:
T. Q. D. Gromard
影响因子:
2.5
作者:
L. Forzani;R. Fraiman;Pamela Llop
通讯作者:
Pamela Llop
影响因子:
1.3
作者:
Phillip A. Ostrand
通讯作者:
Phillip A. Ostrand