Nongaussian subspace learning in the presence of interference
M. Desai, Rami Mangoubi · 2005
We consider the problem of subspace learning in the presence of interference and generalized Gaussian noise, two realistic scenarios for many applications. We also explore learning in the context of a non-Euclidean generalization of the Courant-Fisher minmax characterization. Implications for learned subspace properties in the presence of Laplacian noise are discussed as well.