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

In this paper, we formulate inverse reinforcement learning (IRL) as anexpert-learner interaction whereby the optimal performance intent of an expertor target agent is unknown to a learner agent. The learner observes the statesand controls of the expert and hence seeks to reconstruct the expert's costfunction intent and thus mimics the expert's optimal response. Next, we addnon-cooperative disturbances that seek to disrupt the learning and stability ofthe learner agent. This leads to the formulation of a new interaction we callzero-sum game IRL. We develop a framework to solve the zero-sum game IRLproblem that is a modified extension of RL policy iteration (PI) to allowunknown expert performance intentions to be computed and non-cooperativedisturbances to be rejected. The framework has two parts: a value function andcontrol action update based on an extension of PI, and a cost function updatebased on standard inverse optimal control. Then, we eventually develop anoff-policy IRL algorithm that does not require knowledge of the expert andlearner agent dynamics and performs single-loop learning. Rigorous proofs andanalyses are given. Finally, simulation experiments are presented to show theeffectiveness of the new approach.

Data-Driven Inverse Reinforcement Learning for Expert-Learner Zero-Sum Games