Task Localization, Similarity, and Transfer; Towards a Reinforcement Learning Task Library System
This thesis develops methods of task localization, task similarity discovery, and task transfer for eventual use in a reinforcement learning task library system, which can effectively “learn to learn,” improving its performance as it encounters various tasks over the lifetime of the learning system....
| Main Author: | |
|---|---|
| Format: | Others |
| Published: |
BYU ScholarsArchive
2005
|
| Subjects: | |
| Online Access: | https://scholarsarchive.byu.edu/etd/579 https://scholarsarchive.byu.edu/cgi/viewcontent.cgi?article=1578&context=etd |
| Summary: | This thesis develops methods of task localization, task similarity discovery, and task transfer for eventual use in a reinforcement learning task library system, which can effectively “learn to learn,” improving its performance as it encounters various tasks over the lifetime of the learning system. |
|---|