摘要
生成式对抗网络GAN作为一种生成式模型,受博弈论中二人“零和”博弈的启发,通过对抗学习的方式来训练,可以达到估计数据样本的潜在分布、生成新的数据样本的目的。从GAN的基本原理和实现模型入手,综述了其衍生模型和研究进展,对其在作战体系研究领域的应用进行了分析展望。
Generative Adversarial Networks(GANs)is one type of Generative Models.Inspired by the two-player zero-sum game,GAN is trained with the adversarial learning mechanism.The aim of GAN is to estimate the potential distribution of existing data and generate new data samples from the same distribution.In this paper,the theoretic and implementation are surveyed models and the state of the art of GAN is summarized.Then the further research of the operation SoS oriented application of Generative Adversarial Networks are analyzed.
作者
殷小静
胡晓峰
杨镜宇
YIN Xiao-jing;HU Xiao-feng;YANG Jing-yu(School of Joint Operations,National Defense University,Beijing 100091,China;Unit 96941 of PLA,Beijing 100085,China)
出处
《火力与指挥控制》
CSCD
北大核心
2019年第11期11-15,共5页
Fire Control & Command Control
基金
国家自然科学基金(61703412)
“十三五”装备预先研究基金资助项目(41401030303)
关键词
生成式对抗网络
作战体系
复杂系统建模
样本不足
战场态势认知
generative adversarial networks
operation SoS
complex system modeling
insufficient samples
battlefield situation awareness
