A projection method for under determined optimal experimental designs

Quan Long, Marco Scavino, Raul Tempone, Suojin Wang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A new implementation, based on the Laplace approximation, was developed in (Long, Scavino, Tempone, & Wang 2013) to accelerate the estimation of the post–experimental expected information gains in the model parameters and predictive quantities of interest. A closed–form approximation of the inner integral and the order of the corresponding dominant error term were obtained in the cases where the parameters are determined by the experiment. In this work, we extend that method to the general cases where the model parameters could not be determined completely by the data from the proposed experiments. We carry out the Laplace approximations in the directions orthogonal to the null space of the corresponding Jacobian matrix, so that the information gain (Kullback–Leibler divergence) can be reduced to an integration against the marginal density of the transformed parameters which are not determined by the experiments. Furthermore, the expected information gain can be approximated by an integration over the prior, where the integrand is a function of the projected posterior covariance matrix. To deal with the issue of dimensionality in a complex problem, we use Monte Carlo sampling or sparse quadratures for the integration over the prior probability density function, depending on the regularity of the integrand function. We demonstrate the accuracy, efficiency and robustness of the proposed method via several nonlinear under determined numerical examples.
Original languageEnglish (US)
Title of host publicationSafety, Reliability, Risk and Life-Cycle Performance of Structures and Infrastructures
PublisherInforma UK Limited
Pages2203-2207
Number of pages5
ISBN (Print)9781138000865
DOIs
StatePublished - Jan 13 2014

Fingerprint Dive into the research topics of 'A projection method for under determined optimal experimental designs'. Together they form a unique fingerprint.

Cite this