Meta-Reinforcement Learning with Discrete World Models for Adaptive Load Balancing
Fuente:
arXiv
Gespeichert in:
| 1. Verfasser: | |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917952151879680 |
|---|---|
| author | Redovian, Cameron |
| author_facet | Redovian, Cameron |
| contents | We integrate a meta-reinforcement learning algorithm with the DreamerV3 architecture to improve load balancing in operating systems. This approach enables rapid adaptation to dynamic workloads with minimal retraining, outperforming the Advantage Actor-Critic (A2C) algorithm in standard and adaptive trials. It demonstrates robust resilience to catastrophic forgetting, maintaining high performance under varying workload distributions and sizes. These findings have important implications for optimizing resource management and performance in modern operating systems. By addressing the challenges posed by dynamic and heterogeneous workloads, our approach advances the adaptability and efficiency of reinforcement learning in real-world system management tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_08872 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Meta-Reinforcement Learning with Discrete World Models for Adaptive Load Balancing Redovian, Cameron Machine Learning Artificial Intelligence Operating Systems I.2.6; I.2.8; D.4.1 We integrate a meta-reinforcement learning algorithm with the DreamerV3 architecture to improve load balancing in operating systems. This approach enables rapid adaptation to dynamic workloads with minimal retraining, outperforming the Advantage Actor-Critic (A2C) algorithm in standard and adaptive trials. It demonstrates robust resilience to catastrophic forgetting, maintaining high performance under varying workload distributions and sizes. These findings have important implications for optimizing resource management and performance in modern operating systems. By addressing the challenges posed by dynamic and heterogeneous workloads, our approach advances the adaptability and efficiency of reinforcement learning in real-world system management tasks. |
| title | Meta-Reinforcement Learning with Discrete World Models for Adaptive Load Balancing |
| topic | Machine Learning Artificial Intelligence Operating Systems I.2.6; I.2.8; D.4.1 |
| url | https://arxiv.org/abs/2503.08872 |