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

This paper proposes an exploration method for deep reinforcement learningbased on parameter space noise. Recent studies have experimentally shown thatparameter space noise results in better exploration than the commonly usedaction space noise. Previous methods devised a way to update the diagonalcovariance matrix of a noise distribution and did not consider the direction ofthe noise vector and its correlation. In addition, fast updates of the noisedistribution are required to facilitate policy learning. We propose a methodthat deforms the noise distribution according to the accumulated returns andthe noises that have led to the returns. Moreover, this method switchesisotropic exploration and directional exploration in parameter space withregard to obtained rewards. We validate our exploration strategy in the OpenAIGym continuous environments and modified environments with sparse rewards. Theproposed method achieves results that are competitive with a previous method atbaseline tasks. Moreover, our approach exhibits better performance in sparsereward environments by exploration with the switching strategy.

Switching Isotropic and Directional Exploration with Parameter Space Noise in Deep Reinforcement Learning