eXplainable Artificial Intelligence for RL-based Networking Solutions
By: Yeison Stiven Murcia , Oscar Mauricio Caicedo , Daniela Maria Casas and more
Potential Business Impact:
Shows how smart computer programs make network choices.
Reinforcement Learning (RL) agents have been widely used to improve networking tasks. However, understanding the decisions made by these agents is essential for their broader adoption in networking and network management. To address this, we introduce eXplaNet - a pipeline grounded in explainable artificial intelligence - designed to help networking researchers and practitioners gain deeper insights into the decision-making processes of RL-based solutions. We demonstrate how eXplaNet can be applied to refine a routing solution powered by a Q-learning agent, specifically by improving its reward function. In addition, we discuss the opportunities and challenges of incorporating explainability into RL to better optimize network performance.
Similar Papers
Textual Explanations and Their Evaluations for Reinforcement Learning Policy
Artificial Intelligence
Makes AI explain its decisions like rules.
A Survey of Explainable Reinforcement Learning: Targets, Methods and Needs
Artificial Intelligence
Explains how smart computer programs make decisions.
A Survey on Explainable Deep Reinforcement Learning
Machine Learning (CS)
Makes AI decisions understandable and trustworthy.