Where’s the love?

Our simulation informatics blog hasn’t received much attention recently. I (Paul) have been getting settled in my new position at Colorado School of Mines—learning how to be a professor. David has been busy with his gig at Purdue and his other blog. We’re also both in talks with the folks at SIAM about writing for their upcoming blog; stay tuned for that.

You can stay up-to-date with our latest musings on computational math by following us on Twitter: @DrPaulynomial and @dgleich, respectively.

Model Reduction with MapReduce-enabled Tall and Skinny Singular Value Decomposition

David and I recently submitted a paper on some work we’ve be churning on for a long long time. I’m pretty excited it’s finally submitted. Here’s a link to the preprint on arXiv:

Model Reduction with MapReduce-enabled Tall and Skinny Singular Value Decomposition

Stay tuned for a “Behind the Research” perspective…

Active subspace methods in theory and practice

Qiqi Wang, Eric Dow, and I just submitted our recent work on active subspace methods including a theoretical framework and connections to kriging response surfaces. 

I’m giving talks at the national labs this week telling people about the work. Here’s the link to the arXiv paper:

Active subspace methods in theory and practice: Applications to kriging surfaces

‘Estimating Parameters in Physical Models through Bayesian Inversion: A Complete Example’ by Allmaras, et al

This paper appears in the most recent in issue of SIAM Review, and it is simply fantastic. It is well written and clear, and it judiciously limits the scope to fundamental concepts in statistical inverse problems. I recommend this to anyone interested in such ideas — both students and experienced researchers. In fact, I’m giving a lecture this evening to the UQ class on inverse problems, and I plan to both borrow heavily from its presentation and make it required reading for the class.

Here’s the link.