Information-theoretically optimal sparse PCA

  title={Information-theoretically optimal sparse PCA},
  author={Yash Deshpande and Andrea Montanari},
  journal={2014 IEEE International Symposium on Information Theory},
Sparse Principal Component Analysis (PCA) is a dimensionality reduction technique wherein one seeks a low-rank representation of a data matrix with additional sparsity constraints on the obtained representation. We consider two probabilistic formulations of sparse PCA: a spiked Wigner and spiked Wishart (or spiked covariance) model. We analyze an Approximate Message Passing (AMP) algorithm to estimate the underlying signal and show, in the high dimensional limit, that the AMP estimates are… CONTINUE READING
Highly Cited
This paper has 45 citations. REVIEW CITATIONS
Recent Discussions
This paper has been referenced on Twitter 2 times over the past 90 days. VIEW TWEETS
34 Citations
22 References
Similar Papers


Publications citing this paper.
Showing 1-10 of 34 extracted citations


Publications referenced by this paper.
Showing 1-10 of 22 references

Do semidefinite relaxations really solve sparse pca ?

  • Y. Deshpande, A. Montanari
  • 2013

On finite rank deformations of wigner matrices

  • A. Pizzo, D. Renfrew, A. Soshnikov
  • Annales de l’Institut Henri Poincaré, Probabilit…
  • 2013
1 Excerpt

Sparse pca via covariance threshold

  • A. A. Amini, M. J. Wainwright
  • 2013

Similar Papers

Loading similar papers…