An algorithm for reducing approximate nearest neighbor to approximate near neighbor with O(log n) query time

4Citations
Citations of this article
2Readers
Mendeley users who have this article in their library.
Get full text

Abstract

This paper proposes a new algorithm for reducing Approximate Nearest Neighbor problem to Approximate Near Neighbor problem. The advantage of this algorithm is that it achieves O(log n) query time. As a reduction problem, the query time complexity is the times of invoking the algorithm for Approximate Near Neighbor problem. All former algorithms for the same reduction need polylog(n) query time. A box split method proposed by Vaidya is used in our paper to achieve the O(log n) query time complexity.

Cite

CITATION STYLE

APA

Ma, H., & Li, J. (2018). An algorithm for reducing approximate nearest neighbor to approximate near neighbor with O(log n) query time. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11346 LNCS, pp. 465–479). Springer Verlag. https://doi.org/10.1007/978-3-030-04651-4_31

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free