On the stability of convex sums of rank-1 perturbed matrices.
Thomas LaffeyRobert ShortenFiacre Ó CairbrePublished in: ACC (2002)
Keyphrases
- positive semidefinite matrices
- singular values
- original data
- positive semidefinite
- convex optimization
- low rank
- frobenius norm
- binary matrix
- piecewise linear
- convex sets
- singular value decomposition
- binary matrices
- random variables
- complementarity problems
- low rank approximation
- coefficient matrix
- data matrix
- doubly stochastic
- convex relaxation
- stability analysis
- risk minimization
- linear complementarity problem
- neural network
- globally optimal
- clustering algorithm