Workshop on Statistical Inverse Problems

Statistical Inverse Problems
University of Göttingen
March 23-25, 2006

Organizing Committee
Frank Bauer, Nicolai Bissantz, Thorsten Hohage, Axel Munk


  • Aims and Scope
  • Talks
  • Posters
  • Sponsors
  • Abstracts

Aims and Scope
Inverse Problems is an area of growing interest both for statisticians and numerical analysts since such problems arise naturally in many applications e.g. inin medical imaging, economy, finance, physics, chemistry, biology and industrial research.So far a large part of the research on inverse problems in statisticsand numerics has followed different paths. Whereas a lot of progress has beenachieved on nonlinear deterministic inverse problems over thelast decade, the literature on nonlinear statistical inverse problemsis scarce. In contrast, a variety of sophisticated adaptive techniques forparameter and model selection have been developed in statistics,which do not have a counterpart in deterministic theory.Many questions of fundamental theoretical and practical importancearise in both fields: identifiability,consistency, computation of estimators, and optimality in various forms.Therefore, this workshop intends to establish and strengthen links between research in the statistical and the deterministic inverse problemscommunities.

The workshop covered the following topics:

Methods and Techniques:

  • Algorithmic Aspects of Inverse Problems
  • Bayesian Approaches
  • Minimax Theory
  • Convergence Analysis
  • Iterative Methods for Non-Linear Inverse Problems

Fields of Application:

  • Econometrics
  • Image Reconstruction
  • Deconvolution
  • Medical Applications
  • Technical/Physical/Industrial Applications

Talks download

A. MunkUniversity of GöttingenIntroductionary talk
M. BerteroUniversity of GenovaThe Large Binocular Telescope: A Laboratory for developing Image Reconstruction in Astronomy
L. CavalierUniversité de ProvenceRisk hull method for inverse problems
J. FlorensUniversity of Toulouse IInstrumental regression in partially linear models
N. HengartnerLos Alamos National LaboratiesPassive detection and imaging of nuclear material using cosmic ray muons
J. HorowitzNorthwestern UniversityNonparametric instrumental variables estimation of a quantile
G. JongbloedUniversity of AmsterdamAsymptotic distribution of the MLE in a class of deconvolution models
J. KaipioUniversity of KuopioRecent results in the modelling of approximation errors in inverse problems
P. KimUniversity of GuelphSharp Adaptation for Statistical Inverse Problems on Manifolds
J. LoubesUniversity Paris SudA class of stochastic inverse problems: curves warping
B. MairUniversity of FloridaJoint Emission and Motion Estimation for a Cardiac Cycle in Gated Emission Tomography
E. MammenUniversity of MannheimKernel density estimation for the coefficients in randomcoefficient regression with applications to demand analysis
S. PereverzevJohann Radon Institute for Computational and Applied MathematicsRegularization Algorithms in Learning Theory
M. ReissRuprecht-Karls-Universität HeidelbergCalibration of financial Levy models as inverse problem
F. RuymgaartTexas Tech UniversityFréchet differentiation of functions of operators with application in functional data analysis
E. SomersaloHelsinki University of TechnologyApplications of Bayesian hypermodels
P. StarkUniversity of California Measuring resolution in nonlinear and constrained inverse problems
F. BalabdaouiUniversity of GöttingenEstimation of a convex density: Back to Hampel birds problem
F. BauerUniversity of GöttingenInverse Problems: Strategies to counter noise which exhibits bad behavior
M. PricopUniversity of GöttingenRates of convergence of Tikhonov regularization for nonlinear inverse problems with stochastic noise
N. BissantzUniversity of GöttingenConvergence rates of general regularization methods for statistical inverse problems
L. BoysenUniversity of GöttingenJump reconstruction in certain inverse problems

Posters download

D. CalvettiCase UniversityLarge scale statistical parameter estimation in complex systems with an application to metabolic models
H. HeeseUniversity of GöttingenAn inverse problem in superconductivity
A. HofingerJohann Radon Institute for Computational and Applied MathematicsA new Framework for Assesing Uncertainty in Ill-Posed Problems
O. IvanyshynUniversity of GöttingenNonlinear Integral Equations in Inverse Obstacle Scattering
A. KharytonovKiel UniversityParticle Spectra by the Application of Regularization Methods
M. LangovoyUniversity of GöttingenEfficient tests for the deconvolution hypothesis
F. LenzenUniversity of InnsbruckNon-convex regularization
M. LesoskyUniversity of GuelphStatistical Inverse Problems on the Euclidean Motion Group
T. LevitinaTechnische Universität BraunschweigSampling with Finite Fourier and Hankel Transform Eigenfunctions
C. MarteauUniversite de ProvenceRegularization of inverse problems with noisy operator
Y.M. MarzoukSandia National LaboratoriesStochastic spectral methods for Bayesian inference in inverse problems
M. MeiseUniversität Duisburg-EssenOn Deconvoluting Densities
S.S. Pereverzyev Fraunhofer-Institut für Techno- und WirtschaftsmathematikRegularized Fixed-Point Iteration for Nonlinear Inverse Problems
M.L. RapunUniversidad Complutense de MadridDetecting corrosion using thermal waves
P. SerranhoUniversity of GöttingenA hybrid method for inverse scattering for shape and impedance
H. Weinert/ T. MildenbergerUniversity of DortmundData approximation and inverse problems
V. Zalipaev University of LoughboroughThe evolution-observation scheme in Blagovestchenskii's approach to the 1D inversion

We want to thank all sponsors which make this event possible. In particular these are the "Deutsche Forschungsgemeinschaft" (DFG) and the "Deutsche Akademische Austauschdienst" (DAAD).