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

Model usage is the central challenge of model-based reinforcement learning.Although dynamics model based on deep neural networks provide goodgeneralization for single step prediction, such ability is over exploited whenit is used to predict long horizon trajectories due to compounding errors. Inthis work, we propose a Dyna-style model-based reinforcement learningalgorithm, which we called Maximum Entropy Model Rollouts (MEMR). To eliminatethe compounding errors, we only use our model to generate single-step rollouts.Furthermore, we propose to generate \emph{diverse} model rollouts bynon-uniform sampling of the environment states such that the entropy of themodel rollouts is maximized. We mathematically derived the maximum entropysampling criteria for one data case under Gaussian prior. To accomplish thiscriteria, we propose to utilize a prioritized experience replay. Ourpreliminary experiments in challenging locomotion benchmarks show that ourapproach achieves the same sample efficiency of the best model-basedalgorithms, matches the asymptotic performance of the best model-freealgorithms, and significantly reduces the computation requirements of othermodel-based methods.

Maximum Entropy Model Rollouts: Fast Model Based Policy Optimization without Compounding Errors