On Using Toeplitz and Circulant Matrices for Johnson-Lindenstrauss Transforms
Abstract
The Johnson-Lindenstrauss lemma is one of the corner stone results in dimensionality reduction. It says that given , for any set of vectors , there exists a mapping such that preserves all pairwise distances between vectors in to within if . Much effort has gone into developing fast embedding algorithms, with the Fast Johnson-Lindenstrauss transform of Ailon and Chazelle being one of the most well-known techniques. The current fastest algorithm that yields the optimal dimensions has an embedding time of . An exciting approach towards improving this, due to Hinrichs and Vyb\'iral, is to use a random Toeplitz matrix for the embedding. Using Fast Fourier Transform, the embedding of a vector can then be computed in time. The big question is of course whether dimensions suffice for this technique. If so, this would end a decades long quest to obtain faster and faster Johnson-Lindenstrauss transforms. The current best analysis of the embedding of Hinrichs and Vyb\'iral shows that dimensions suffices. The main result of this paper, is a proof that this analysis unfortunately cannot be tightened any further, i.e., there exists a set of vectors requiring for the Toeplitz approach to work.
Keywords
Cite
@article{arxiv.1706.10110,
title = {On Using Toeplitz and Circulant Matrices for Johnson-Lindenstrauss Transforms},
author = {Casper Benjamin Freksen and Kasper Green Larsen},
journal= {arXiv preprint arXiv:1706.10110},
year = {2017}
}