On solving a Stochastic Shortest-Path Markov Decision Process as probabilistic inference
Previous work on planning as active inference addresses finite horizon problems and solutions valid for online planning. We propose solving the general Stochastic Shortest-Path Markov Decision Process (SSP MDP) as probabilistic inference. Furthermore, we discuss online and offline methods for planni...
| Main Authors: | , , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
Springer
2022
|
| _version_ | 1797109039725805568 |
|---|---|
| author | Baioumy, M Lacerda, B Duckworth, P Hawes, N |
| author_facet | Baioumy, M Lacerda, B Duckworth, P Hawes, N |
| author_sort | Baioumy, M |
| collection | OXFORD |
| description | Previous work on planning as active inference addresses finite horizon problems and solutions valid for online planning. We propose solving the general Stochastic Shortest-Path Markov Decision Process (SSP MDP) as probabilistic inference. Furthermore, we discuss online and offline methods for planning under uncertainty. In an SSP MDP, the horizon is indefinite and unknown a priori. SSP MDPs generalize finite and infinite horizon MDPs and are widely used in the artificial intelligence community. Additionally, we highlight some of the differences between solving an MDP using dynamic programming approaches widely used in the artificial intelligence community and approaches used in the active inference community. F |
| first_indexed | 2024-03-07T07:36:29Z |
| format | Conference item |
| id | oxford-uuid:70e31dd9-0431-435e-a90b-6619000a4160 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T07:36:29Z |
| publishDate | 2022 |
| publisher | Springer |
| record_format | dspace |
| spelling | oxford-uuid:70e31dd9-0431-435e-a90b-6619000a41602023-03-14T12:05:16ZOn solving a Stochastic Shortest-Path Markov Decision Process as probabilistic inferenceConference itemhttp://purl.org/coar/resource_type/c_5794uuid:70e31dd9-0431-435e-a90b-6619000a4160EnglishSymplectic ElementsSpringer2022Baioumy, MLacerda, BDuckworth, PHawes, NPrevious work on planning as active inference addresses finite horizon problems and solutions valid for online planning. We propose solving the general Stochastic Shortest-Path Markov Decision Process (SSP MDP) as probabilistic inference. Furthermore, we discuss online and offline methods for planning under uncertainty. In an SSP MDP, the horizon is indefinite and unknown a priori. SSP MDPs generalize finite and infinite horizon MDPs and are widely used in the artificial intelligence community. Additionally, we highlight some of the differences between solving an MDP using dynamic programming approaches widely used in the artificial intelligence community and approaches used in the active inference community. F |
| spellingShingle | Baioumy, M Lacerda, B Duckworth, P Hawes, N On solving a Stochastic Shortest-Path Markov Decision Process as probabilistic inference |
| title | On solving a Stochastic Shortest-Path Markov Decision Process as probabilistic inference |
| title_full | On solving a Stochastic Shortest-Path Markov Decision Process as probabilistic inference |
| title_fullStr | On solving a Stochastic Shortest-Path Markov Decision Process as probabilistic inference |
| title_full_unstemmed | On solving a Stochastic Shortest-Path Markov Decision Process as probabilistic inference |
| title_short | On solving a Stochastic Shortest-Path Markov Decision Process as probabilistic inference |
| title_sort | on solving a stochastic shortest path markov decision process as probabilistic inference |
| work_keys_str_mv | AT baioumym onsolvingastochasticshortestpathmarkovdecisionprocessasprobabilisticinference AT lacerdab onsolvingastochasticshortestpathmarkovdecisionprocessasprobabilisticinference AT duckworthp onsolvingastochasticshortestpathmarkovdecisionprocessasprobabilisticinference AT hawesn onsolvingastochasticshortestpathmarkovdecisionprocessasprobabilisticinference |