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 introduce a new count-based optimistic exploration algorithm forReinforcement Learning (RL) that is feasible in environments withhigh-dimensional state-action spaces. The success of RL algorithms in thesedomains depends crucially on generalisation from limited training experience.Function approximation techniques enable RL agents to generalise in order toestimate the value of unvisited states, but at present few methods enablegeneralisation regarding uncertainty. This has prevented the combination ofscalable RL algorithms with efficient exploration strategies that drive theagent to reduce its uncertainty. We present a new method for computing ageneralised state visit-count, which allows the agent to estimate theuncertainty associated with any state. Our \phi-pseudocount achievesgeneralisation by exploiting same feature representation of the state spacethat is used for value function approximation. States that have less frequentlyobserved features are deemed more uncertain. The \phi-Exploration-Bonusalgorithm rewards the agent for exploring in feature space rather than in theuntransformed state space. The method is simpler and less computationallyexpensive than some previous proposals, and achieves near state-of-the-artresults on high-dimensional RL benchmarks.

Count-Based Exploration in Feature Space for Reinforcement Learning