English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Bayesian optimization (BO) is a flexible and powerful framework that issuitable for computationally expensive simulation-based applications andguarantees statistical convergence to the global optimum. While remaining asone of the most popular optimization methods, its capability is hindered by thesize of data, the dimensionality of the considered problem, and the nature ofsequential optimization. These scalability issues are intertwined with eachother and must be tackled simultaneously. In this work, we propose theScalable$^3$-BO framework, which employs sparse GP as the underlying surrogatemodel to scope with Big Data and is equipped with a random embedding toefficiently optimize high-dimensional problems with low effectivedimensionality. The Scalable$^3$-BO framework is further leveraged withasynchronous parallelization feature, which fully exploits the computationalresource on HPC within a computational budget. As a result, the proposedScalable$^3$-BO framework is scalable in three independent perspectives: withrespect to data size, dimensionality, and computational resource on HPC. Thegoal of this work is to push the frontiers of BO beyond its well-knownscalability issues and minimize the wall-clock waiting time for optimizinghigh-dimensional computationally expensive applications. We demonstrate thecapability of Scalable$^3$-BO with 1 million data points, 10,000-dimensionalproblems, with 20 concurrent workers in an HPC environment.

Scalable3-BO: Big Data meets HPC - A scalable asynchronous parallel high-dimensional Bayesian optimization framework on supercomputers