|
Vol.11, No.1, February 2022. ISSN: 2217-8309 eISSN: 2217-8333
TEM Journal
TECHNOLOGY, EDUCATION, MANAGEMENT, INFORMATICS Association for Information Communication Technology Education and Science |
Segmented Actor-Critic-Advantage Architecture for Reinforcement Learning Tasks
Martin Kaloev, Georgi Krastev
© 2022 Martin Kaloev, published by UIKTEN. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. (CC BY-NC-ND 4.0)
Citation Information: TEM Journal. Volume 11, Issue 1, Pages 219-224, ISSN 2217-8309, DOI: 10.18421/TEM111-27, February 2022.
Received: 04 January 2022. Revised: 05 February 2022.
Abstract:
The article focuses on experiments with a multi module neural networks type of architecture for neuron-like machine used in reinforcing learning. This type of architecture can be used to solve complex robotic or policy optimization tasks and allows segmented storage of trained memory. Such technique speeds up the training process compared to existing actor-critical algorithms.
Keywords – Reinforcement learning, Q-learning, Actor-critic algorithm, Neuron-like machine architecture. |
|
----------------------------------------------------------------------------------------------------------- ----------------------------------------------------------------------------------------------------------- |