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 investigate the effect of including domain knowledge about a roboticsystem's causal relations when generating explanations. To this end, we comparetwo methods from explainable artificial intelligence, the popular KernelSHAPand the recent causal SHAP, on a deep neural network trained using deepreinforcement learning on the task of controlling a lever using a roboticmanipulator. A primary disadvantage of KernelSHAP is that its explanationsrepresent only the features' direct effects on a model's output, notconsidering the indirect effects a feature can have on the output by affectingother features. Causal SHAP uses a partial causal ordering to alterKernelSHAP's sampling procedure to incorporate these indirect effects. Thispartial causal ordering defines the causal relations between the features, andwe specify this using domain knowledge about the lever control task. We showthat enabling an explanation method to account for indirect effects andincorporating some domain knowledge can lead to explanations that better agreewith human intuition. This is especially favorable for a real-world roboticstask, where there is considerable causality at play, and in addition, therequired domain knowledge is often handily available.

Causal versus Marginal Shapley Values for Robotic Lever Manipulation Controlled using Deep Reinforcement Learning