Improving k-NN graph accuracy using local intrinsic dimensionality
Document Type
Conference Proceeding
Publication Date
1-1-2017
Abstract
The k-nearest neighbor (k-NN) graph is an important data structure for many data mining and machine learning applications. The accuracy of k-NN graphs depends on the object feature vectors, which are usually represented in high-dimensional spaces. Selecting the most important features is essential for providing compact object representations and for improving the graph accuracy. Having a compact feature vector can reduce the storage space and the computational complexity of search and learning tasks. In this paper, we propose NNWID-Descent, a similarity graph construction method that utilizes the NNF-Descent framework while integrating a new feature selection criterion, Support-Weighted Intrinsic Dimensionality, that estimates the contribution of each feature to the overall intrinsic dimensionality. Through extensive experiments on various datasets, we show that NNWID-Descent allows a significant amount of local feature vector sparsification while still preserving a reasonable level of graph accuracy.
Identifier
85031303591 (Scopus)
ISBN
[9783319684734]
Publication Title
Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics
External Full Text Location
https://doi.org/10.1007/978-3-319-68474-1_8
e-ISSN
16113349
ISSN
03029743
First Page
110
Last Page
124
Volume
10609 LNCS
Grant
25240036
Fund Ref
Japan Society for the Promotion of Science
Recommended Citation
Houle, Michael E.; Oria, Vincent; and Wali, Arwa M., "Improving k-NN graph accuracy using local intrinsic dimensionality" (2017). Faculty Publications. 9970.
https://digitalcommons.njit.edu/fac_pubs/9970
