Abstract
This study presents an enhancement of the Market-Driven Multi-agent Reinforcement Learning Hierarchical Neuro-Fuzzy Model (MA-RL-HNFP-MD) by incorporating Deep Reinforcement Learning (DRL) techniques. The modified framework, referred to as the Deep Reinforcement Learning Market-Driven Hierarchical Neuro-Fuzzy Model (DRL-MD-HNFP), utilizes advanced neural network architectures to improve learning efficiency and scalability in complex multi-agent environments. Experimental results demonstrate significant performance improvements compared to the original MA-RL-HNFP-MD model, particularly in reducing task completion times and optimizing resource allocation. We validated the proposed model in benchmark scenarios, including the pursuit game and robotic soccer simulations, where it exhibited superior adaptability and coordination among agents. These findings highlight the potential of DRL-based approaches to enhance decision-making and coordination in multi-agent systems, providing valuable insights for applications in dynamic and resource-intensive environments.
| Original language | English |
|---|---|
| Title of host publication | Anais do XVII Congresso Brasileiro de Inteligência Computacional |
| Editors | Frederico Gadelha Guimarães, Danton Diego Ferreira, Guilherme de Alencar Barreto |
| Number of pages | 5 |
| ISBN (Electronic) | 978-65-978625-0-4 |
| DOIs | |
| State | Published - 2025 |
| Event | XVII Congresso Brasileiro de Inteligência Computacional - Belo Horizonte, Brazil Duration: 27 Oct 2025 → 30 Oct 2025 Conference number: 17 https://sbia.org.br/eventos/cbic_2025/ |
Conference
| Conference | XVII Congresso Brasileiro de Inteligência Computacional |
|---|---|
| Country/Territory | Brazil |
| City | Belo Horizonte |
| Period | 27/10/25 → 30/10/25 |
| Internet address |
UN SDGs
This output contributes to the following UN Sustainable Development Goals (SDGs)
-
SDG 9 Industry, Innovation, and Infrastructure
-
SDG 17 Partnerships for the Goals
Fingerprint
Dive into the research topics of 'Enhancing market-driven multi-agent systems with deep reinforcement learning: A hierarchical neuro-fuzzy approach'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver