Show simple item record

dc.contributor.authorLeikanger, Per Roald
dc.date.accessioned2022-04-06T09:56:50Z
dc.date.available2022-04-06T09:56:50Z
dc.date.issued2021-07-19
dc.description.abstractNavigating the world is a fundamental ability for any living entity. Accomplishing the same degree of freedom in technology has proven to be difficult. The brain is the only known mechanism capable of voluntary navigation, making neuroscience our best source of inspiration toward autonomy. Assuming that state representation is key, we explore the difference in how the brain and the machine represent the navigational state. Where Reinforcement Learning (RL) requires a monolithic state representation in accordance with the Markov property, Neural Representation of Euclidean Space (NRES) reflects navigational state via distributed activation patterns. We show how NRES-Oriented RL (neoRL) agents are possible before verifying our theoretical findings by experiments. Ultimately, neoRL agents are capable of behavior synthesis across state spaces – allowing for decomposition of the problem into smaller spaces, alleviating the curse of dimensionality.en_US
dc.identifier.citationLeikanger PR: Decomposing the Prediction Problem; Autonomous Navigation by neoRL Agents. In: Cejkova J, Holler, Soros, Witkowski O. ALIFE 2021: Proceedings of the Artificial Life Conference 2021, 2021. MIT Pressen_US
dc.identifier.cristinIDFRIDAID 2009892
dc.identifier.doi10.1162/isal_a_00444
dc.identifier.issn2693-1508
dc.identifier.urihttps://hdl.handle.net/10037/24721
dc.language.isoengen_US
dc.publisherMIT Pressen_US
dc.relation.ispartofLeikanger, P.R. (2022). Autonomous Navigation in (the Animal and) the Machine. (Doctoral thesis). <a href=https://hdl.handle.net/10037/25518>https://hdl.handle.net/10037/25518</a>.
dc.rights.accessRightsopenAccessen_US
dc.rights.holderCopyright 2021 Massachusetts Institute of Technologyen_US
dc.titleDecomposing the Prediction Problem; Autonomous Navigation by neoRL Agentsen_US
dc.type.versionpublishedVersionen_US
dc.typeJournal articleen_US
dc.typeTidsskriftartikkelen_US


File(s) in this item

Thumbnail

This item appears in the following collection(s)

Show simple item record