Expected Improvements for the Asynchronous Parallel Global Optimization of Expensive Functions : Potentials and Challenges

Abstract : Sequential sampling strategies based on Gaussian processes are now widely used for the optimization of problems involving costly simulations. But Gaussian processes can also generate parallel optimiza- tion strategies. We focus here on a new, parameter free, parallel expected improvement criterion for asynchronous optimization. An estimation of the criterion, which mixes Monte Carlo sampling and analytical bounds, is proposed. Logarithmic speed-ups are measured on 1 and 9 dimensional functions.
Complete list of metadatas

https://hal-emse.ccsd.cnrs.fr/emse-00686504
Contributor : Florent Breuil <>
Submitted on : Tuesday, April 10, 2012 - 2:34:22 PM
Last modification on : Thursday, October 17, 2019 - 12:36:13 PM

Links full text

Identifiers

Citation

Janis Janusevskis, Rodolphe Le Riche, David Ginsbourger, Ramunas Girdziusas. Expected Improvements for the Asynchronous Parallel Global Optimization of Expensive Functions : Potentials and Challenges. Lecture Notes in Computer Science, Springer, 2012, Learning and Intelligent Optimization, pp. 413-418. ⟨10.1007/978-3-642-34413-8_37⟩. ⟨emse-00686504⟩

Share

Metrics

Record views

212