From Roman Vershynin's presentation at SamPTA  entitled: Sampling and high-dimensional convex geometry, here are two slides that caught my attention:
and most importantly
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.