Intrinsic dimension estimation: Advances and open problems

被引：101
作者：
Camastra, Francesco [1 ]
Staiano, Antonino [1 ]
机构：
[1] Univ Naples Parthenope, Dept Sci & Technol, I-80143 Naples, Italy
来源：
INFORMATION SCIENCES | 2016年 / 328卷
关键词：
Intrinsic dimension; Curse of dimensionality; Maximum likelihood; Correlation dimension; Dimensionality reduction; NONLINEAR PCA; MANIFOLDS; LIMITATIONS; REDUCTION; SELECTION; NETWORKS;
D O I：
10.1016/j.ins.2015.08.029
中图分类号：
TP [自动化技术、计算机技术];
学科分类号：
0812 ;
摘要：
Dimensionality reduction methods are preprocessing techniques used for coping with high dimensionality. They have the aim of projecting the original data set of dimensionality N, without information loss, onto a lower M-dimensional submanifold. Since the value of M is unknown, techniques that allow knowing in advance the value of M, called intrinsic dimension (ID), are quite useful. The aim of the paper is to review state-of-the-art of the methods of ID estimation, underlining the recent advances and the open problems. (C) 2015 Elsevier Inc. All rights reserved.
引用
收藏
页码：26 / 41
页数：16
相关论文
共 104 条
  • [1] [Anonymous], 1993, An introduction to the bootstrap
  • [2] [Anonymous], 1977, Fractals, Form, Chance, and Dimension
  • [3] [Anonymous], 1988, CHAOS DYNAMICAL SYST
  • [4] [Anonymous], 1993, NETWORKS CHAOS STAT
  • [5] [Anonymous], 2004, Advances in neural information processing systems, DOI DOI 10.5555/2976040.2976138
  • [6] [Anonymous], 1961, Adaptive Control Processes: a Guided Tour, DOI DOI 10.1515/9781400874668
  • [7] [Anonymous], COLLECTED WORKS LEJ
  • [8] [Anonymous], 2003, Advances in Neural Information Processing Systems 15, DOI DOI 10.1109/34.682189
  • [9] [Anonymous], P 18 INT C ART INT S
  • [10] [Anonymous], 1982, PATTERN RECOGN
12345678910