ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1607.06017
14
50

Doubly Accelerated Methods for Faster CCA and Generalized Eigendecomposition

20 July 2016
Zeyuan Allen-Zhu
Yuanzhi Li
ArXivPDFHTML
Abstract

We study kkk-GenEV, the problem of finding the top kkk generalized eigenvectors, and kkk-CCA, the problem of finding the top kkk vectors in canonical-correlation analysis. We propose algorithms LazyEV\mathtt{LazyEV}LazyEV and LazyCCA\mathtt{LazyCCA}LazyCCA to solve the two problems with running times linearly dependent on the input size and on kkk. Furthermore, our algorithms are DOUBLY-ACCELERATED: our running times depend only on the square root of the matrix condition number, and on the square root of the eigengap. This is the first such result for both kkk-GenEV or kkk-CCA. We also provide the first gap-free results, which provide running times that depend on 1/ε1/\sqrt{\varepsilon}1/ε​ rather than the eigengap.

View on arXiv
Comments on this paper