Download | - View final version: Dynamic programming with incomplete information to overcomenavigational uncertainty in POMDPs (PDF, 1.6 MiB)
|
---|
Link | https://caiac.pubpub.org/pub/qdmqsaj7 |
---|
Author | Search for: Beeler, Chris1; Search for: Li, Xinkai Li; Search for: Bellinger, Colin1; Search for: Crowley, Mark; Search for: Fraser, Maia; Search for: Tamblyn, Isaac |
---|
Affiliation | - National Research Council of Canada. Digital Technologies
|
---|
Format | Text, Article |
---|
Conference | The 37th Canadian Conference on Artificial Intelligence (Canadian AI 2024), May 27-31, 2024, Guelph, Ontario, Canada |
---|
Subject | dynamic programming; partially observable; Markov decision processes; risk management; controlled sensing |
---|
Abstract | Using a generalizable novel nautical navigation environment, we show how dynamic programming can be used when only incomplete information about a partially observed Markov decision process (POMDP) is known. By incorporating uncertainty into our model, we show that navigation policies can be constructed that maintain safety, outperforming the baseline performance of traditional dynamic programming for Markov decision processes (MDPs). Adding in controlled sensing methods, we show that these policies can also lower measurement costs at the same time. |
---|
Publication date | 2024-05-27 |
---|
Publisher | Canadian Artificial Intelligence Association |
---|
Licence | |
---|
In | |
---|
Language | English |
---|
Peer reviewed | Yes |
---|
Export citation | Export as RIS |
---|
Report a correction | Report a correction (opens in a new tab) |
---|
Record identifier | aa2e9fc6-dbcb-4a14-a604-9d0951bb7183 |
---|
Record created | 2024-06-13 |
---|
Record modified | 2024-06-14 |
---|