Hui Huang; Haber, Eldad; Horesh, Lior; Jin Keun Seo
August 2012
Inverse Problems & Imaging;Aug2012, Vol. 6 Issue 3, p447
Academic Journal
We address the problem of prior matrix estimation for the solution of â„“1-regularized ill-posed inverse problems. From a Bayesian viewpoint, we show that such a matrix can be regarded as an in uence matrix in a multivariate â„“1-Laplace density function. Assuming a training set is given, the prior matrix design problem is cast as a maximum likelihood term with an additional sparsity-inducing term. This formulation results in an unconstrained yet nonconvex optimization problem. Memory requirements as well as computation of the nonlinear, nonsmooth sub-gradient equations are prohibitive for large-scale problems. Thus, we introduce an iterative algorithm to design efficient priors for such large problems. We further demonstrate that the solutions of ill-posed inverse problems by incorporation of â„“1-regularization using the learned prior matrix perform generally better than commonly used regularization techniques where the prior matrix is chosen a-priori.


Related Articles

  • Stochastic optimization with structured distributions: The case of Bayesian nets. Gaivoronski, A.; Stella, F. // Annals of Operations Research;1998, Vol. 81 Issue 1-4, p189 

    In this paper, the authors consider interworking between statistical procedures for recovering the distribution of random parameters from observations and stochastic programming techniques, in particular stochastic gradient (quasigradient) methods. The proposed problem formulation is based upon...

  • Bayesian Wavelet Based Signal and Image Separation. Ichir, Mahieddine M.; Mohammad-Djafari, Ali // AIP Conference Proceedings;2004, Vol. 707 Issue 1, p417 

    In this contribution, we consider the problem of blind source separation in a Bayesian estimation framework. The wavelet representation allows us to assign an adequate prior distribution to the wavelet coefficients of the sources. MCMC algorithms are implemented to test the validity of the...

  • A BAYESIAN ALGORITHM FOR GLOBAL OPTIMIZATION. Betró, B.; Rotondi, R. // Annals of Operations Research;1984, Vol. 1 Issue 1-4, p111 

    A crucial step in global optimization algorithms based on random sampling in the search domain is decision about the achievement of a prescribed accuracy. In order to overcome the difficulties related to such a decision, the Bayesian Nonparametric Approach has been introduced. The aim of this...

  • VALUATION-BASED SYSTEMS FOR BAYESIAN DECISION ANALYSIS. Shenoy, Prakash P. // Operations Research;May/Jun92, Vol. 40 Issue 3, p463 

    This paper proposes a new method for representing and solving Bayesian decision problems. The representation is called a valuation-based system and has some similarities to influence diagrams. However, unlike influence diagrams which emphasize conditional independence among random variables,...

  • A Bayesian Classifier Learning Algorithm Based on Optimization Model. Sanyang Liu; Mingmin Zhu; Youlong Yang // Mathematical Problems in Engineering;2013, p1 

    Naive Bayes classifier is a simple and effective classification method, but its attribute independence assumption makes it unable to express the dependence among attributes and affects its classification performance. In this paper, we summarize the existing improved algorithms and propose a...

  • A Simple Bayesian Procedure for Estimation in a Conjoint Model. Cattin, Philippe; Gelfand, Alan E.; Danes, Jeffrey // Journal of Marketing Research (JMR);Feb1983, Vol. 20 Issue 1, p29 

    The authors propose a simple Bayesian approach which combines self-explicated data with conjoint data for estimating individual-level conjoint models. Analytical results show that, with typical conjoint data, improvement may be expected over the estimation and prediction results obtained with...

  • ON BAYESIAN GROUP SEQUENTIAL SAMPLING PROCEDURES. Rieder, Ulrich; Wentges, Paul // Annals of Operations Research;1991, Vol. 32 Issue 1-4, p189 

    A general group sequential statistical decision model is investigated. Using a Bayesian approach and Bayesian dynamic programming, we examine structural properties of the cost functions and of optimal sampling procedures. In particular, we consider variable-sample-size-sequential probability...

  • A Framework for Initial Experimental Design in the Presence of Competing Prior Knowledge. VASTÓLA, JUSTIN T.; JYE-CHYI LU; CASCIATO, MICHAEL J.; HESS, DENNIS W.; GROVER, MARTHA A. // Journal of Quality Technology;Oct2013, Vol. 45 Issue 4, p301 

    An initial experimental design criteria is proposed for optimization problems by combining multiple, potentially competing, sources of prior information--engineering models, expert opinion, and data from past experimentation on similar, nonidentical systems. By leveraging prior information,...

  • An adaptive stochastic global optimization algorithm for one-dimensional functions. Locatelli, Marco; Schoen, Fabio // Annals of Operations Research;1995, Vol. 58 Issue 1-4, p263 

    In this paper a new algorithm is proposed, based upon the idea of modeling the objective function of a global optimization problem as a sample path from a Wiener process. Unlike previous work in this field, in the proposed model the parameter of the Wiener process is considered as a random...


Read the Article


Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics