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

Learning robot controllers by minimizing a black-box objective cost usingBayesian optimization (BO) can be time-consuming and challenging. It is veryoften the case that some roll-outs result in failure behaviors, causingpremature experiment detention. In such cases, the designer is forced to decideon heuristic cost penalties because the acquired data is often scarce, or notcomparable with that of the stable policies. To overcome this, we propose aBayesian model that captures exactly what we know about the cost of unstablecontrollers prior to data collection: Nothing, except that it should be asomewhat large number. The resulting Bayesian model, approximated with aGaussian process, predicts high cost values in regions where failures arelikely to occur. In this way, the model guides the BO exploration towardregions of stability. We demonstrate the benefits of the proposed model inseveral illustrative and statistical synthetic benchmarks, and also inexperiments on a real robotic platform. In addition, we propose andexperimentally validate a new BO method to account for unknown constraints.Such method is an extension of Max-Value Entropy Search, a recentinformation-theoretic method, to solve unconstrained global optimizationproblems.

Classified Regression for Bayesian Optimization: Robot Learning with Unknown Penalties