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

Scientific experiments are usually expensive due to complex experimentalpreparation and processing. Experimental design is therefore involved with thetask of finding the optimal experimental input that results in the desirableoutput by using as few experiments as possible. Experimenters can often acquirethe knowledge about the location of the global optimum. However, they do notknow how to exploit this knowledge to accelerate experimental design. In thispaper, we adopt the technique of Bayesian optimization for experimental designsince Bayesian optimization has established itself as an efficient tool foroptimizing expensive black-box functions. Again, it is unknown how toincorporate the expert prior knowledge about the global optimum into Bayesianoptimization process. To address it, we represent the expert knowledge aboutthe global optimum via placing a prior distribution on it and we then deriveits posterior distribution. An efficient Bayesian optimization approach hasbeen proposed via posterior sampling on the posterior distribution of theglobal optimum. We theoretically analyze the convergence of the proposedalgorithm and discuss the robustness of incorporating expert prior. We evaluatethe efficiency of our algorithm by optimizing synthetic functions and tuninghyperparameters of classifiers along with a real-world experiment on thesynthesis of short polymer fiber. The results clearly demonstrate theadvantages of our proposed method.

Incorporating Expert Prior Knowledge into Experimental Design via Posterior Sampling