Scalable Algorithms for Inverse Problems With High-Dimensional Parameter Spaces
Fairbanks, Hillary · 2025
Read Scalable Algorithms for Inverse Problems With High-Dimensional Parameter Spaces by Fairbanks, Hillary (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/doe-osti-2584729-scalable-algorithms-for-inverse-problems.txt.
Description
Hillary Fairbanks of Lawrence Livermore National Laboratory wrote report LLNL-TR-2005705 in June 2025 on scalable algorithms for inverse problems. Inverse problems infer unknown values from observed data, and they grow hard when the unknown set is huge and the links to the data are nonlinear. The report focuses on Bayesian inference, which frames the unknown values as a probability distribution.
Fairbanks explains why Markov chain Monte Carlo sampling becomes too costly as problem size grows, since each step may need a costly partial differential equation solve. She reviews methods that cut that cost, including multilevel sampling, delayed acceptance using cheaper model approximations, and approaches that need far fewer equation solves. The goal is practical Bayesian inference for large, high-dimensional systems.
Catalog record
| CREATOR | Fairbanks, Hillary |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 9 KB |
| LICENSE | Public Domain |
| RIGHTS | PUBLIC DOMAIN |
| SOURCE | DOE OSTI |