Sparse Surrogate Model Construction via Compressive Sensing for High-Dimensional Complex Models

Abstract

For complex models with a large number of input parameters,
surrogate model construction is challenged by insufficient
model simulation data as well as by a prohibitively large number of
parameters controlling the surrogate. Bayesian sparse learning approaches are
implemented in order to detect sparse-basis expansions that best
capture the model outputs. We enhanced the Bayesian compressive
sensing approach with adaptive basis growth and with a data-driven,
piecewise surrogate construction.


Date
Jul 11, 2013
Location
San Diego, CA