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 searching for a generalizable representation of temporally extended tasks,we spot two necessary constituents: the utility needs to be non-Markovian totransfer temporal relations invariant to a probability shift, the utility alsoneeds to be lifted to abstract out specific grounding objects. In this work, westudy learning such utility from human demonstrations. While inversereinforcement learning (IRL) has been accepted as a general framework ofutility learning, its fundamental formulation is one concrete Markov DecisionProcess. Thus the learned reward function does not specify the taskindependently of the environment. Going beyond that, we define a domain ofgeneralization that spans a set of planning problems following a schema. Wehence propose a new quest, Generalized Inverse Planning, for utility learningin this domain. We further outline a computational framework, Maximum EntropyInverse Planning (MEIP), that learns non-Markovian utility and associatedconcepts in a generative manner. The learned utility and concepts form a taskrepresentation that generalizes regardless of probability shift or structuralchange. Seeing that the proposed generalization problem has not been widelystudied yet, we carefully define an evaluation protocol, with which weillustrate the effectiveness of MEIP on two proof-of-concept domains and onechallenging task: learning to fold from demonstrations.

Generalized Inverse Planning: Learning Lifted non-Markovian Utility for Generalizable Task Representation