Bayesian Compressive Sensing Framework for Sparse Representations of High-Dimensional Models


Surrogate construction for high-dimensional models is challenged in two major ways:
obtaining sufficient training model simulations becomes prohibitively
expensive, and non-adaptive basis selection rules lead to excessively large basis sets.
We enhanced select state-of-the-art tools from statistical learning to build efficient sparse
surrogate representations, with quantified uncertainty, for high-dimensional complex models.
Specifically, Bayesian compressive sensing techniques are supplemented by iterative basis growth
and weighted regularization. Application to an 80-dimensional
climate land model shows promising results.

Apr 1, 2014
Savannah, GA