Sparsity and Overcomplete Data Representation
May 17th, 2005
The Israel Statistics Asociation Annual Meeting, Tel-Aviv, Israel.

A recent trend in signal, image, and data analysis is the use of overcomplete linear transforms that lead to a sparse description of the processed data. This new breed of methods is more difficult to use, but they are much more effective in applications such as data compression and solution of inverse problems. In this talk we present a wide angle view to this recent path of research.