Advances in MCMC and related sampling methods for large-scale inverse problems

来源 :第八届工业与应用数学国际大会 | 被引量 : 0次 | 上传用户:btly540205390
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
  Inverse problems convert indirect measurements into useful characterizations of the parameters of a physical system.Parameters are typically related to indirect measurements by a system of partial differential equations(PDEs),which are complicated and expensive to evaluate.
其他文献
  This mini-symposium will present an overview of recent advances on the control of PDEs and stochastic PDEs,as well as their challenging applicative issues f
会议
  This talk presents some progress and comments on time optimal control problems of heat equations.It focuses on the bang-bang property.We will present a new
会议
  In this paper,we report three level implicit method of high accuracy schemes of O(k4+h8),O(k6+h8)and O(k6+h10)for the numerical solution of fourth order non
会议
  A mathematical modeling for stored corn proposed by Lopes et al.[Lopes D.de C.et al.2006.Aeration simulation of stored grain under variable air ambient cond
会议
  We present in this work a numerical study of the square-root formulation which is a stabilization method for computing highly elastic flows of viscoelastic
会议
  A typical two-state repairable system is considered in this paper.To analyze the instantaneous reliability index of the system,we describe the system as an
会议
  A Chebyshev-Legendre Pseudo-Spectral method is applied for solving Kupershmidt equation.In time direction we used a leapfrog scheme,while Chebyshev-Legendre
会议
  In Bayesian inference,the uncertainty of parameters of a physical system is characterized by a probability distribution.Numerical algorithms are typically u
会议
  Ensemble Prediction Systems(EPS)are used to quantify the uncertainty of weather predictions,by launching predictions with perturbed initial values.
会议
  We present an approach(based on Bayes' formula and ratio estimates)to apply QMC and multilevel MC methods for the computation of posterior expectations of f
会议