Speaker
Dr
Mikko Vehkaperä
(KTH and Aalto University)
Description
The sparse representation problem of recovering an N
dimensional sparse vector x from M < N linear observations
y = Dx given dictionary D is considered. The standard
approach is to let the elements of the dictionary be
independent and identically distributed (IID) zero-mean
Gaussian and minimize the l1 norm of x under the
constraint y = Dx. In this talk, we discuss the replica
analysis of l1-reconstruction when the dictionary is a
concatenation D = [O1 O2 ... OL] of L independent
orthogonal random matrices. The results indicate that
under some conditions, the compression threshold for such
D is the same as for the IID case.
This is recent joint work with Y Kabashima and others partly available as [arXiv:1204.4065]
This is recent joint work with Y Kabashima and others partly available as [arXiv:1204.4065]
Primary author
Dr
Mikko Vehkaperä
(KTH and Aalto University)
