Abstract: Deep reinforcement learning (DRL) algorithms have shown impressive results in various applications, but nonstationary environments, such as varying operating conditions and external ...