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

We present a new algorithm ASEBO for optimizing high-dimensional blackboxfunctions. ASEBO adapts to the geometry of the function and learns optimal setsof sensing directions, which are used to probe it, on-the-fly. It addresses theexploration-exploitation trade-off of blackbox optimization with expensiveblackbox queries by continuously learning the bias of the lower-dimensionalmodel used to approximate gradients of smoothings of the function viacompressed sensing and contextual bandits methods. To obtain this model, itleverages techniques from the emerging theory of active subspaces in the novelES blackbox optimization context. As a result, ASEBO learns the dynamicallychanging intrinsic dimensionality of the gradient space and adapts to thehardness of different stages of the optimization without external supervision.Consequently, it leads to more sample-efficient blackbox optimization thanstate-of-the-art algorithms. We provide theoretical results and test ASEBOadvantages over other methods empirically by evaluating it on the set ofreinforcement learning policy optimization tasks as well as functions from therecently open-sourced Nevergrad library.

From Complexity to Simplicity: Adaptive ES-Active Subspaces for Blackbox Optimization