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

With the increasing presence of robotic systems and human-robot environmentsin today's society, understanding the reasoning behind actions taken by a robotis becoming more important. To increase this understanding, users are providedwith explanations as to why a specific action was taken. Among other effects,these explanations improve the trust of users in their robotic partners. Oneoption for creating these explanations is an introspection-based approach whichcan be used in conjunction with reinforcement learning agents to provideprobabilities of success. These can in turn be used to reason about the actionstaken by the agent in a human-understandable fashion. In this work, thisintrospection-based approach is developed and evaluated further on the basis ofan episodic and a non-episodic robotics simulation task. Furthermore, anadditional normalization step to the Q-values is proposed, which enables theusage of the introspection-based approach on negative and comparatively smallQ-values. Results obtained show the viability of introspection for episodicrobotics tasks and, additionally, that the introspection-based approach can beused to generate explanations for the actions taken in a non-episodic roboticsenvironment as well.

Introspection-based Explainable Reinforcement Learning in Episodic and Non-episodic Scenarios