Codes and Expansions (CodEx) Seminar

Karin Schnass (Universität Innsbruck)
Random submatrices with non-uniformly distributed supports

We will motivate why it is useful to look at random sub-dictionaries (subframes) where each atom is not drawn with the same probability but some atoms are more likely.

We will then provide conditions under which such sub-dictionaries are well-conditioned with high probability, sketch the proof and show the crux of the proof in more detail.

Time permitting we will then discuss how these results can be used to decide what a good sensing matrix for compressed sensing is and how we can precondition an existing one.

Joint work with Simon Ruetz.