A Study of Asynchronous Budgeted Optimization
NIPS 2012 Workshop on Bayesian Optimization & Decision Making 2012
Rodolphe Le Riche, Jānis Januševskis, Ramunas Girdziusas

Budgeted optimization algorithms based on Gaussian processes have attracted a lot of attention as a way to deal with computationally expensive cost functions. Recently, parallel versions of these algorithms have further improved their ability to address the computation cost bottleneck. This article focuses on those algorithms that maximize the multi-point expected improvement criterion. Synchronous and asynchronous parallel versions of such algorithms are compared. It is shown that asynchronous algorithms are slower than the synchronous ones iteration-wise. In terms of wall clock time however and contrarily to synchronous algorithms, asynchronous implementations benefit from a near-linear speed-up up to the order of 100 computing nodes. This speed-up is bounded by the maximization of the expected improvement time which, by becoming a critical time limitation, may change the way researchers look at budgeted optimization methods.


Atslēgas vārdi
Bayesian Optimization, Computer Grids, Clud optimization
Hipersaite
http://javad-azimi.com/nips2012ws/Schedule.html

Le Riche, R., Januševskis, J., Girdziusas, R. A Study of Asynchronous Budgeted Optimization. No: NIPS 2012 Workshop on Bayesian Optimization & Decision Making, Amerikas savienotās valstis, Lake Tahoe, 3.-9. decembris, 2012. Internets: 2012, 1.-5.lpp.

Publikācijas valoda
English (en)
RTU Zinātniskā bibliotēka.
E-pasts: uzzinas@rtu.lv; Tālr: +371 28399196