Structured Low-Rank Approximation as Optimization on a Grassmann Manifold

Many data modeling problems can be posed and solved as a structured low-rank approximation problem. Using the variable projection approach, the problem is reformulated as optimization on a Grassmann m