05459nam 2200661Ia 450 991083073740332120210209181003.01-282-16506-297866121650610-470-61119-70-470-39382-3(CKB)2550000000005884(EBL)477672(SSID)ssj0000335046(PQKBManifestationID)11272597(PQKBTitleCode)TC0000335046(PQKBWorkID)10271369(PQKB)10114410(MiAaPQ)EBC477672(MiAaPQ)EBC4037004(PPN)19071316X(OCoLC)521032626(EXLCZ)99255000000000588420071119d2008 uy 0engur|n|---|||||txtccrBayesian approach to inverse problems[electronic resource] /edited by Jerome IdierLondon ISTE ;Hoboken, NJ John Wileyc20081 online resource (383 p.)Digital signal and image processing series. ;v.35Description based upon print version of record.1-84821-032-9 Includes bibliographical references and index.Bayesian Approach to Inverse Problems; Table of Contents; Introduction; Part I. Fundamental Problems and Tools; Chapter 1. Inverse Problems, Ill-posed Problems; 1.1. Introduction; 1.2. Basic example; 1.3. Ill-posed problem; 1.3.1. Case of discrete data; 1.3.2. Continuous case; 1.4. Generalized inversion; 1.4.1. Pseudo-solutions; 1.4.2. Generalized solutions; 1.4.3. Example; 1.5. Discretization and conditioning; 1.6. Conclusion; 1.7. Bibliography; Chapter 2. Main Approaches to the Regularization of Ill-posed Problems; 2.1. Regularization; 2.1.1. Dimensionality control2.1.1.1. Truncated singular value decomposition2.1.1.2. Change of discretization; 2.1.1.3. Iterative methods; 2.1.2. Minimization of a composite criterion; 2.1.2.1. Euclidian distances; 2.1.2.2. Roughness measures; 2.1.2.3. Non-quadratic penalization; 2.1.2.4. Kullback pseudo-distance; 2.2. Criterion descent methods; 2.2.1. Criterion minimization for inversion; 2.2.2. The quadratic case; 2.2.2.1. Non-iterative techniques; 2.2.2.2. Iterative techniques; 2.2.3. The convex case; 2.2.4. General case; 2.3. Choice of regularization coefficient; 2.3.1. Residual error energy control2.3.2. "L-curve" method2.3.3. Cross-validation; 2.4. Bibliography; Chapter 3. Inversion within the Probabilistic Framework; 3.1. Inversion and inference; 3.2. Statistical inference; 3.2.1. Noise law and direct distribution for data; 3.2.2. Maximum likelihood estimation; 3.3. Bayesian approach to inversion; 3.4. Links with deterministic methods; 3.5. Choice of hyperparameters; 3.6. A priori model; 3.7. Choice of criteria; 3.8. The linear, Gaussian case; 3.8.1. Statistical properties of the solution; 3.8.2. Calculation of marginal likelihood; 3.8.3. Wiener filtering; 3.9. BibliographyPart II. DeconvolutionChapter 4. Inverse Filtering and Other Linear Methods; 4.1. Introduction; 4.2. Continuous-time deconvolution; 4.2.1. Inverse filtering; 4.2.2. Wiener filtering; 4.3. Discretization of the problem; 4.3.1. Choice of a quadrature method; 4.3.2. Structure of observation matrix H; 4.3.3. Usual boundary conditions; 4.3.4. Problem conditioning; 4.3.4.1. Case of the circulant matrix; 4.3.4.2. Case of the Toeplitz matrix; 4.3.4.3. Opposition between resolution and conditioning; 4.3.5. Generalized inversion; 4.4. Batch deconvolution; 4.4.1. Preliminary choices4.4.2. Matrix form of the estimate4.4.3. Hunt's method (periodic boundary hypothesis); 4.4.4. Exact inversion methods in the stationary case; 4.4.5. Case of non-stationary signals; 4.4.6. Results and discussion on examples; 4.4.6.1. Compromise between bias and variance in 1D deconvolution; 4.4.6.2. Results for 2D processing; 4.5. Recursive deconvolution; 4.5.1. Kalman filtering; 4.5.2. Degenerate state model and recursive least squares; 4.5.3. Autoregressive state model; 4.5.3.1. Initialization; 4.5.3.2. Criterion minimized by Kalman smoother; 4.5.3.3. Example of result4.5.4. Fast Kalman filteringMany scientific, medical or engineering problems raise the issue of recovering some physical quantities from indirect measurements; for instance, detecting or quantifying flaws or cracks within a material from acoustic or electromagnetic measurements at its surface is an essential problem of non-destructive evaluation. The concept of inverse problems precisely originates from the idea of inverting the laws of physics to recover a quantity of interest from measurable data.Unfortunately, most inverse problems are ill-posed, which means that precise and stable solutions are not easy to deviseDigital signal and image processing series.Inverse problems (Differential equations)Bayesian statistical decision theoryInverse problems (Differential equations)Bayesian statistical decision theory.515/.357519.542Idier Jérôme1593281MiAaPQMiAaPQMiAaPQBOOK9910830737403321Bayesian approach to inverse problems3913340UNINA