Optimal Solutions for Sparse Principal Component Analysis

  • Authors: A. d'Aspremont, F. Bach, L. El Ghaoui.

  • Status: Journal of Machine Learning Research, 9(Jul):1269–1294, 2008.

  • Abstract: Given a sample covariance matrix, we examine the problem of maximizing the variance explained by a linear combination of the input variables while constraining the number of nonzero coefficients in this combination. This is known as sparse principal component analysis and has a wide array of applications in machine learning and engineering. We formulate a new semidefinite relaxation to this problem and derive a greedy algorithm that computes a full set of good solutions for all target numbers of non zero coefficients, with total complexity O(n^3), where n is the number of variables. We then use the same relaxation to derive sufficient conditions for global optimality of a solution, which can be tested in O(n^3) per pattern. We discuss applications in subset selection and sparse recovery and show on artificial examples and biological data that our algorithm does provide globally optimal solutions in many cases.

  • Bibtex reference:

@article{ABE:08a,
	Author = {A. d'Aspremont, F. Bach and L. {El Ghaoui}},
	Journal = {Journal of Machine Learning Research},
	Month = {July},
	Pages = {1269--1294},
	Title = {Optimal Solutions for Sparse Principal Component Analysis},
	Volume = {9},
	Year = {2008}}