Comparison of DDPG and TD3 Algorithms in a Walker2D Scenario

DOI

10.2991/978-94-6463-370-2_17How to use a DOI?

Keywords

Reinforcement Learning; DDPG; TD3

Abstract

Reinforcement learning has emerged as a powerful approach for tackling complex continuous control tasks across various domains. This paper presents an extensive comparative analysis of two prominent reinforcement learning algorithms: the Deep Deterministic Policy Gradient (DDPG) algorithm and its advanced counterpart, the Twin-Delayed DDPG (TD3) algorithm. The primary focus is on evaluating the performance and effectiveness of these algorithms within the realm of locomotion control, a domain with substantial real-world implications. This study centers around the Walker2D problem, a challenging locomotion control task available in the OpenAI Gym environment. Walker2D presents a compelling testbed for assessing the practicality of reinforcement learning algorithms in contexts such as robotics, autonomous systems, and physical control. By conducting a detailed examination of DDPG and TD3, the author aims to shed light on their strengths and weaknesses in continuous control scenarios. Beyond academic interest, this research has significant real-world relevance. Mastery of continuous control tasks holds immense promise for applications ranging from robotics and automation to healthcare and beyond. In essence, this study bridges the gap between theoretical advancements in reinforcement learning and their practical implications in solving real-world challenges. By providing a comprehensive evaluation of these algorithms in the demanding context of locomotion control, this work contributes to the broader understanding of reinforcement learning’s potential to drive innovation and efficiency in various domains.

Copyright

© 2024 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Xinrui Shen
PY  - 2024
DA  - 2024/02/14
TI  - Comparison of DDPG and TD3 Algorithms in a Walker2D Scenario
BT  - Proceedings of the 2023 International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2023)
PB  - Atlantis Press
SP  - 148
EP  - 155
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-370-2_17
DO  - 10.2991/978-94-6463-370-2_17
ID  - Shen2024
ER  -