English

Multiscale Shrinkage and L\'evy Processes

Machine Learning 2014-01-14 v1

Abstract

A new shrinkage-based construction is developed for a compressible vector xRn\boldsymbol{x}\in\mathbb{R}^n, for cases in which the components of \xv\xv are naturally associated with a tree structure. Important examples are when \xv\xv corresponds to the coefficients of a wavelet or block-DCT representation of data. The method we consider in detail, and for which numerical results are presented, is based on increments of a gamma process. However, we demonstrate that the general framework is appropriate for many other types of shrinkage priors, all within the L\'{e}vy process family, with the gamma process a special case. Bayesian inference is carried out by approximating the posterior with samples from an MCMC algorithm, as well as by constructing a heuristic variational approximation to the posterior. We also consider expectation-maximization (EM) for a MAP (point) solution. State-of-the-art results are manifested for compressive sensing and denoising applications, the latter with spiky (non-Gaussian) noise.

Keywords

Cite

@article{arxiv.1401.2497,
  title  = {Multiscale Shrinkage and L\'evy Processes},
  author = {Xin Yuan and Vinayak Rao and Shaobo Han and Lawrence Carin},
  journal= {arXiv preprint arXiv:1401.2497},
  year   = {2014}
}

Comments

11 pages, 5 figures

R2 v1 2026-06-22T02:43:15.702Z