.png){kind=icon}
Return
Anytime Automatic Algorithm Selection for the Multi-Agent Path Finding Problem
.png){kind=icon}
OA
AI
DOI：10.1109/ACCESS.2024.3395495.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In this study, we propose and develop a Machine Learning-based metasolver for the Multi-Agent Path Finding (MAPF) problem, with the aim of selecting the most suitable solver based on the specific characteristics of the problem and a user-provided time constraint. The approach aims to improve the performance of the best-performing solver on average and approximate the performance of a perfect selector. To achieve this, a comprehensive and diverse dataset was compiled, and state-of-the-art algorithms were selected and modified to efficiently handle the time constraint. Also, relevant features were identified, and a precise and robust Machine Learning model was constructed using the XGBoost algorithm. The model was evaluated and compared against other state-of-the-art methods. The results demonstrate that the proposed approach is effective and consistent, outperforming the Single Best Solver and approximating the performance of the Virtual Best Solver.
Keywords:
Algorithm design and analysis
Multi-agent systems
Path planning
Time factors
Approximation algorithms
Performance evaluation
Algorithm selection
multi-agent path finding
machine learning
Journal
IF:
3.6
Papers:
9.8W
Citations:
29.4W
Organization
Cited Papers
ShareSaveA Survey on Visual Navigation for Artificial Agents With Deep Reinforcement Learning.png)
IEEE ACCESS
OAAI
IEEE ACCESSIF3.6
OAAIShareSaveShareSaveShareSaveShareSaveShareSave
View more