English

LazySVD: Even Faster SVD Decomposition Yet Without Agonizing Pain

Numerical Analysis 2017-01-24 v2 Data Structures and Algorithms Machine Learning Optimization and Control Machine Learning

Abstract

We study kk-SVD that is to obtain the first kk singular vectors of a matrix AA. Recently, a few breakthroughs have been discovered on kk-SVD: Musco and Musco [1] proved the first gap-free convergence result using the block Krylov method, Shamir [2] discovered the first variance-reduction stochastic method, and Bhojanapalli et al. [3] provided the fastest O(nnz(A)+poly(1/ε))O(\mathsf{nnz}(A) + \mathsf{poly}(1/\varepsilon))-time algorithm using alternating minimization. In this paper, we put forward a new and simple LazySVD framework to improve the above breakthroughs. This framework leads to a faster gap-free method outperforming [1], and the first accelerated and stochastic method outperforming [2]. In the O(nnz(A)+poly(1/ε))O(\mathsf{nnz}(A) + \mathsf{poly}(1/\varepsilon)) running-time regime, LazySVD outperforms [3] in certain parameter regimes without even using alternating minimization.

Keywords

Cite

@article{arxiv.1607.03463,
  title  = {LazySVD: Even Faster SVD Decomposition Yet Without Agonizing Pain},
  author = {Zeyuan Allen-Zhu and Yuanzhi Li},
  journal= {arXiv preprint arXiv:1607.03463},
  year   = {2017}
}

Comments

first circulated on May 20, 2016; this newer version improves writing