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 data-efficient framework for solving sequential decision-makingproblems which exploits the combination of reinforcement learning (RL) andlatent variable generative models. The framework, called GenRL, trains deeppolicies by introducing an action latent variable such that the feed-forwardpolicy search can be divided into two parts: (i) training a sub-policy thatoutputs a distribution over the action latent variable given a state of thesystem, and (ii) unsupervised training of a generative model that outputs asequence of motor actions conditioned on the latent action variable. GenRLenables safe exploration and alleviates the data-inefficiency problem as itexploits prior knowledge about valid sequences of motor actions. Moreover, weprovide a set of measures for evaluation of generative models such that we areable to predict the performance of the RL policy training prior to the actualtraining on a physical robot. We experimentally determine the characteristicsof generative models that have most influence on the performance of the finalpolicy training on two robotics tasks: shooting a hockey puck and throwing abasketball. Furthermore, we empirically demonstrate that GenRL is the onlymethod which can safely and efficiently solve the robotics tasks compared totwo state-of-the-art RL methods.

Training and Evaluation of Deep Policies using Reinforcement Learning and Generative Models