Enhanced EVA with Threshold Limit of Similarity

In recent years, there has been a growing demand for efficient learning algorithms to apply reinforcing learning to real-world tasks. The algorithm, called EVA, utilized non-parametric Q-values without using neural networks to improve efficiency of learning. We pointed out the problems with EVA, and...

Full description

Saved in:
Bibliographic Details
Published in2021 IEEE 10th Global Conference on Consumer Electronics (GCCE) pp. 329 - 331
Main Authors Tsuneda, Toi, Imade, Kuniyasu, Shintani, Kosuke, Yamane, Satoshi
Format Conference Proceeding
LanguageEnglish
Published IEEE 12.10.2021
Subjects
Conferences
Consumer electronics
Deep learning
Episodic memory
Machine learning
Neural networks
Reinforcement learning
Task analysis
Online AccessGet full text

Cover

More Information
Summary:In recent years, there has been a growing demand for efficient learning algorithms to apply reinforcing learning to real-world tasks. The algorithm, called EVA, utilized non-parametric Q-values without using neural networks to improve efficiency of learning. We pointed out the problems with EVA, and proposed the improved algorithm with even better performance. We conducted comparative experiments on several tasks to assess the efficacy of our proposed methodology.
DOI:10.1109/GCCE53005.2021.9621981