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

Deep neuroevolution and deep reinforcement learning (deep RL) algorithms aretwo popular approaches to policy search. The former is widely applicable andrather stable, but suffers from low sample efficiency. By contrast, the latteris more sample efficient, but the most sample efficient variants are alsorather unstable and highly sensitive to hyper-parameter setting. So far, thesefamilies of methods have mostly been compared as competing tools. However, anemerging approach consists in combining them so as to get the best of bothworlds. Two previously existing combinations use either an ad hoc evolutionaryalgorithm or a goal exploration process together with the Deep DeterministicPolicy Gradient (DDPG) algorithm, a sample efficient off-policy deep RLalgorithm. In this paper, we propose a different combination scheme using thesimple cross-entropy method (CEM) and Twin Delayed Deep Deterministic policygradient (td3), another off-policy deep RL algorithm which improves over ddpg.We evaluate the resulting method, cem-rl, on a set of benchmarks classicallyused in deep RL. We show that cem-rl benefits from several advantages over itscompetitors and offers a satisfactory trade-off between performance and sampleefficiency.

CEM-RL: Combining evolutionary and gradient-based methods for policy search