Abstract | ||
---|---|---|
Catastrophic forgetting is of special importance in reinforcement learning, as the data distribution is generally non-stationary over time. We study and compare several pseudorehearsal approaches for Q-learning with function approximation in a pole balancing task. We have found that pseudorehearsal seems to assist learning even in such very simple problems, given proper initialization of the rehearsal parameters. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1007/978-3-319-59394-4_18 | AGENT AND MULTI-AGENT SYSTEMS: TECHNOLOGY AND APPLICATIONS |
Keywords | Field | DocType |
Reinforcement learning,Rehearsal,Pseudorehearsal,Catastrophic forgetting | Forgetting,Function approximation,Computer science,Pole balancing,Bellman equation,Artificial intelligence,Initialization,Machine learning,Reinforcement learning | Journal |
Volume | ISSN | Citations |
74 | 2190-3018 | 3 |
PageRank | References | Authors |
0.54 | 10 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Vladimir Marochko | 1 | 3 | 0.54 |
Leonard Johard | 2 | 11 | 4.91 |
Manuel Mazzara | 3 | 493 | 64.05 |