Exploiting Multi-Task Learning to Achieve Effective Transfer Deep Reinforcement Learning in Elastic Optical Networks
Xiaoliang Chen, Roberto Proietti, Che-Yu Liu, Zuqing Zhu, S. J. Ben Yoo · 2020
We propose a multi-task-learning-aided knowledge transferring approach for effective and scalable deep reinforcement learning in EONs. Case studies with RMSA show that this approach can achieve ∼ 4 × learning time reduction and ∼ 17 . 7% lower blocking probability.