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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Deep Reinforcement Learning (Deep RL) and Evolutionary Algorithms (EA) aretwo major paradigms of policy optimization with distinct learning principles,i.e., gradient-based v.s. gradient-free. An appealing research direction isintegrating Deep RL and EA to devise new methods by fusing their complementaryadvantages. However, existing works on combining Deep RL and EA have two commondrawbacks: 1) the RL agent and EA agents learn their policies individually,neglecting efficient sharing of useful common knowledge; 2) parameter-levelpolicy optimization guarantees no semantic level of behavior evolution for theEA side. In this paper, we propose Evolutionary Reinforcement Learning withTwo-scale State Representation and Policy Representation (ERL-Re$^2$), a novelsolution to the aforementioned two drawbacks. The key idea of ERL-Re$^2$ istwo-scale representation: all EA and RL policies share the same nonlinear staterepresentation while maintaining individual} linear policy representations. Thestate representation conveys expressive common features of the environmentlearned by all the agents collectively; the linear policy representationprovides a favorable space for efficient policy optimization, where novelbehavior-level crossover and mutation operations can be performed. Moreover,the linear policy representation allows convenient generalization of policyfitness with the help of the Policy-extended Value Function Approximator(PeVFA), further improving the sample efficiency of fitness estimation. Theexperiments on a range of continuous control tasks show that ERL-Re$^2$consistently outperforms advanced baselines and achieves the State Of The Art(SOTA). Our code is available on https://github.com/yeshenpy/ERL-Re2.

ERL-Re$^2$: Efficient Evolutionary Reinforcement Learning with Shared State Representation and Individual Policy Representation