Genet
Zhengxu Xia, Yajie Zhou, Francis Y. Yan, Junchen Jiang
Abstract
As deep reinforcement learning (RL) showcases its strengths in networking, its pitfalls are also coming to the public's attention. Training on a wide range of network environments leads to suboptimal performance, whereas training on a narrow distribution of environments results in poor generalization.
Topics & Concepts
Reinforcement learningGeneralizationComputer scienceArtificial intelligenceRange (aeronautics)EngineeringMathematicsMathematical analysisAerospace engineeringNeural Networks and Reservoir ComputingImage and Video Quality AssessmentData Stream Mining Techniques