High-Rank Matrix Completion and Subspace Clustering with Missing Data Brian Eriksson∗ Boston University and University of Wisconsin - Madison Laura Balzano∗ University of Wisconsin - Madison Robert Nowak University of Wisconsin - Madison December 2011 Abstract This paper considers the problem of completing a matrix with many missing entries under the assumption that the columns of the matrix belong to a union of multiple low-rank subspaces.
- rank matrix completion
- deviation of an incomplete vector norm with respect to the incoherence
- subspaces
- subspace
- rank
- columns
- high probability
- seed
- matrix
- probability