Automatic Optimization of Electro-Optic Frequency Comb Based on Deep Reinforcement Learning

We propose a novel automatically optimized electro-optic frequency comb (AO-EOFC) based on deep reinforcement learning, which utilizes a deep Q-learning network algorithm to replace blind and inefficient manual optimization in traditional ways.