Interpretable Reinforcement Learning for Load Balancing using Kolmogorov-Arnold Networks
This work addresses interpretability issues in RL for network control, which is incremental as it applies a novel method to a known bottleneck in a specific domain.
The paper tackled the problem of lack of interpretability in reinforcement learning for network load balancing by using Kolmogorov-Arnold Networks (KAN) to extract controller equations, resulting in improved network performance with maximized throughput utility and minimized loss and delay.
Reinforcement learning (RL) has been increasingly applied to network control problems, such as load balancing. However, existing RL approaches often suffer from lack of interpretability and difficulty in extracting controller equations. In this paper, we propose the use of Kolmogorov-Arnold Networks (KAN) for interpretable RL in network control. We employ a PPO agent with a 1-layer actor KAN model and an MLP Critic network to learn load balancing policies that maximise throughput utility, minimize loss as well as delay. Our approach allows us to extract controller equations from the learned neural networks, providing insights into the decision-making process. We evaluate our approach using different reward functions demonstrating its effectiveness in improving network performance while providing interpretable policies.