Investigating Transfer Learning in Noisy Environments: A Study of Predecessor and Successor Features in Spatial Learning Using a T-Maze

Citations

WEB OF SCIENCE

2
Citations

SCOPUS

2

초록

In this study, we investigate the adaptability of artificial agents within a noisy T-maze that use Markov decision processes (MDPs) and successor feature (SF) and predecessor feature (PF) learning algorithms. Our focus is on quantifying how varying the hyperparameters, specifically the reward learning rate (alpha r) and the eligibility trace decay rate (lambda), can enhance their adaptability. Adaptation is evaluated by analyzing the hyperparameters of cumulative reward, step length, adaptation rate, and adaptation step length and the relationships between them using Spearman's correlation tests and linear regression. Our findings reveal that an alpha r of 0.9 consistently yields superior adaptation across all metrics at a noise level of 0.05. However, the optimal setting for lambda varies by metric and context. In discussing these results, we emphasize the critical role of hyperparameter optimization in refining the performance and transfer learning efficacy of learning algorithms. This research advances our understanding of the functionality of PF and SF algorithms, particularly in navigating the inherent uncertainty of transfer learning tasks. By offering insights into the optimal hyperparameter configurations, this study contributes to the development of more adaptive and robust learning algorithms, paving the way for future explorations in artificial intelligence and neuroscience.

키워드

reinforcement learning; T-maze transfer learning; hyperparameter tuning; robustness; predecessor features; noisy observation
제목
Investigating Transfer Learning in Noisy Environments: A Study of Predecessor and Successor Features in Spatial Learning Using a T-Maze
저자
Seo, Incheol; Lee, Hyunsu
DOI
10.3390/s24196419
발행일
2024-10
유형
Article
저널명
Sensors
권
24
호
19