Stochastic Geometry and Random Matrix Theory in Compressed Sensing

Jared Tanner
University of Edinburgh

Stochastic Geometry and Random Matrix Theory have provided tools for the rigorous
analysis of recovery conditions and sampling theorems in compressed sensing. These
perspectives will be reviewed and recent results presented. In particular, the following
two items will be presented: the best known bounds on restricted isometry constants of
large rectangular matrices, as and a full characterisation of compressed sensing for
vectors with bound constraints.

Presentation (PDF File)

Back to Workshop II: Numerical Methods for Continuous Optimization