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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Gaussian Process bandit optimization has emerged as a powerful tool foroptimizing noisy black box functions. One example in machine learning ishyper-parameter optimization where each evaluation of the target functionrequires training a model which may involve days or even weeks of computation.Most methods for this so-called "Bayesian optimization" only allow sequentialexploration of the parameter space. However, it is often desirable to proposebatches or sets of parameter values to explore simultaneously, especially whenthere are large parallel processing facilities at our disposal. Batch methodsrequire modeling the interaction between the different evaluations in thebatch, which can be expensive in complex scenarios. In this paper, we propose anew approach for parallelizing Bayesian optimization by modeling the diversityof a batch via Determinantal point processes (DPPs) whose kernels are learnedautomatically. This allows us to generalize a previous result as well as provebetter regret bounds based on DPP sampling. Our experiments on a variety ofsynthetic and real-world robotics and hyper-parameter optimization tasksindicate that our DPP-based methods, especially those based on DPP sampling,outperform state-of-the-art methods.

Batched Gaussian Process Bandit Optimization via Determinantal Point Processes