基于机器学习的作战体系能力特征指标挖掘
|
摘要
针对当前作战体系能力特征指标挖掘存在的两个困难:作战数据生成和挖掘方法选择,提出"先采用仿真试验床生成作战数据,再利用机器学习挖掘特征指标"的方法。研究了2种基于机器学习的特征指标挖掘方法:基于网络聚合的方法,依据基础指标的相关性进行社团划分,利用主成分分析法得到特征指标,应用该方法挖掘防空能力的特征指标;基于集成学习的方法,利用装袋法生成测试数据集和CART决策树训练模型,利用主成分分析法得到特征指标,应用该方法挖掘防突入能力的特征指标。
Aiming at the two difficulties in characteristic index digging of combat system of systems(CSoS),namely operation data generation and digging method selection, this paper proposes a new digging method,that is, using the simulation testbed to generate operation data, then adopting the machine learning to dig characteristic index. Two methods of characteristic index digging based on machine learning are researched:(1)the method based on network convergence, divides the communities for fundamental indexes based on their relationship, and obtains the characteristic indexes by principal component analysis(PCA); this method is applied to dig the characteristic indexes of air defense ability.(2) the method based on ensemble learning,generates test data by bagging, trains model by CART decision trees, and obtains the characteristic indexes by PCA; this method is applied to dig the characteristic indexes of air defense breakthrough ability.
Aiming at the two difficulties in characteristic index digging of combat system of systems(CSoS),namely operation data generation and digging method selection, this paper proposes a new digging method,that is, using the simulation testbed to generate operation data, then adopting the machine learning to dig characteristic index. Two methods of characteristic index digging based on machine learning are researched:(1)the method based on network convergence, divides the communities for fundamental indexes based on their relationship, and obtains the characteristic indexes by principal component analysis(PCA); this method is applied to dig the characteristic indexes of air defense ability.(2) the method based on ensemble learning,generates test data by bagging, trains model by CART decision trees, and obtains the characteristic indexes by PCA; this method is applied to dig the characteristic indexes of air defense breakthrough ability.
引文
[1]李凤雷,卢昊,宋闯,等.智能化战争与无人系统技术的发展[J].无人系统技术,2018,1(2):14-23.Li Fenglei,Lu Hao,Song Chuang,et al.Development of Intelligent Warfare and Unmanned System Technology[J].Unmanned Systems Technology,2018,1(2):14-23.
[2]伍文峰,胡晓峰.基于大数据的网络化作战体系能力评估框架[J].军事运筹与系统工程,2016,30(2):26-32.Wu Wengfeng,Hu Xiaofeng.Capability Evaluation Framework of Networked Operations System Based on Big Data[J].Military Operations Research and Systems Engineering,2016,30(2):26-32.
[3]丁剑飞.武器装备体系效能指标体系构建方法研究[D].北京:国防大学,2017.Ding Jianfei.Research on Construction Method of Weapon Equipment System Effectiveness Index System[D].Beijing:PLA National Denfense University,2017.
[4]张昱,杨镜宇,李仁见,等.面向武器装备体系仿真的五环聚焦实验框架[J].系统仿真学报,2014,26(11):2541-2545.Zhang Yu,Yang Jingyu,Li Renjian,et al.5-Loop Framework of Making and Converging Experiments for Simulation of Weapon System of Systems[J].Journal of System Simulation,2014,26(11):2541-2545.
[5]Silver D,Huang A,Maddison C J,et al.Mastering the game of Go with deep neural networks and tree search[J].Nature(S1476-4687),2016,529(7587):484-489.
[6]姚晓毅,郭圣明,胡晓峰,等.基于MINE-FNA组合算法的防空体系指标网构建方法[J].指挥控制与仿真,2016,38(1):94-100.Yao Xiaoyi,Guo Shengming,Hu Xiaofeng,et al.Indices Network Construction for Air Defense System of Systems Based on MINE-FNA Assembled Algorithm[J].Command Control&Simulation,2016,38(1):94-100.
[7]郭圣明,贺筱媛,吴琳,等.基于强制稀疏自编码神经网络的防空体系能力回溯分析方法研究[J].中国科学:信息科学,2018,48(7):824-840.Guo Shengming,He Xiaoyuan,Wu Lin,et al.Backtracking Analysis Approach for Effectiveness of Air Defense Operation System of Systems Based on Force-sparsed Stacked-autoencoding Neural Networks[J].SCIENTIA SINICA Informationis,2018,48(7):824-840.
[8]蒋杭进.最大信息系数及其在脑网络分析中的应用[D].武汉:中国科学院研究生院(武汉物理与数学研究所),2013.Jiang Hangjin.Maximal Information Coefficient and Its Annlication to Brain Network Analysis[D].Wuhan:Graduate University of Chinese Academy of Sciences(Wuhan Institute of Physics and Mathematics),2013.
[9]Girvan M,Newman M E J.Community structure in social and biological networks[J].Proceedings of the National Academy of Sciences(S1091-6490),2002,99(12):7821-7826.
[10]魏旭.基于主成分分析的特征融合及其应用[D].成都:电子科技大学,2008.Wei Xu.Feature Fusion Based on Principal Component Analysis and Tts Application[D].Chengdu:University of Electronic Science and Technology,2008.
[11]Dietterich T G.Machine learning research:Four current directions[J].AI Magazine(S0738-4602),1997,18(4):97-136.
[12]Zhou Z H,Wu J X,Tang W.Ensembling neural networks:Many could be better than all[J].Artificial Intelligence(S0004-3702),2002,137(1/2):239-263.
[13]Breiman L.Bagging predictors[J].Machine Learning(S0885-6125),1996,24(2):123-140.
[14]Breiman L.Classification and regression trees[M].New York:Chapman&Hall/CRC,1984.
[2]伍文峰,胡晓峰.基于大数据的网络化作战体系能力评估框架[J].军事运筹与系统工程,2016,30(2):26-32.Wu Wengfeng,Hu Xiaofeng.Capability Evaluation Framework of Networked Operations System Based on Big Data[J].Military Operations Research and Systems Engineering,2016,30(2):26-32.
[3]丁剑飞.武器装备体系效能指标体系构建方法研究[D].北京:国防大学,2017.Ding Jianfei.Research on Construction Method of Weapon Equipment System Effectiveness Index System[D].Beijing:PLA National Denfense University,2017.
[4]张昱,杨镜宇,李仁见,等.面向武器装备体系仿真的五环聚焦实验框架[J].系统仿真学报,2014,26(11):2541-2545.Zhang Yu,Yang Jingyu,Li Renjian,et al.5-Loop Framework of Making and Converging Experiments for Simulation of Weapon System of Systems[J].Journal of System Simulation,2014,26(11):2541-2545.
[5]Silver D,Huang A,Maddison C J,et al.Mastering the game of Go with deep neural networks and tree search[J].Nature(S1476-4687),2016,529(7587):484-489.
[6]姚晓毅,郭圣明,胡晓峰,等.基于MINE-FNA组合算法的防空体系指标网构建方法[J].指挥控制与仿真,2016,38(1):94-100.Yao Xiaoyi,Guo Shengming,Hu Xiaofeng,et al.Indices Network Construction for Air Defense System of Systems Based on MINE-FNA Assembled Algorithm[J].Command Control&Simulation,2016,38(1):94-100.
[7]郭圣明,贺筱媛,吴琳,等.基于强制稀疏自编码神经网络的防空体系能力回溯分析方法研究[J].中国科学:信息科学,2018,48(7):824-840.Guo Shengming,He Xiaoyuan,Wu Lin,et al.Backtracking Analysis Approach for Effectiveness of Air Defense Operation System of Systems Based on Force-sparsed Stacked-autoencoding Neural Networks[J].SCIENTIA SINICA Informationis,2018,48(7):824-840.
[8]蒋杭进.最大信息系数及其在脑网络分析中的应用[D].武汉:中国科学院研究生院(武汉物理与数学研究所),2013.Jiang Hangjin.Maximal Information Coefficient and Its Annlication to Brain Network Analysis[D].Wuhan:Graduate University of Chinese Academy of Sciences(Wuhan Institute of Physics and Mathematics),2013.
[9]Girvan M,Newman M E J.Community structure in social and biological networks[J].Proceedings of the National Academy of Sciences(S1091-6490),2002,99(12):7821-7826.
[10]魏旭.基于主成分分析的特征融合及其应用[D].成都:电子科技大学,2008.Wei Xu.Feature Fusion Based on Principal Component Analysis and Tts Application[D].Chengdu:University of Electronic Science and Technology,2008.
[11]Dietterich T G.Machine learning research:Four current directions[J].AI Magazine(S0738-4602),1997,18(4):97-136.
[12]Zhou Z H,Wu J X,Tang W.Ensembling neural networks:Many could be better than all[J].Artificial Intelligence(S0004-3702),2002,137(1/2):239-263.
[13]Breiman L.Bagging predictors[J].Machine Learning(S0885-6125),1996,24(2):123-140.
[14]Breiman L.Classification and regression trees[M].New York:Chapman&Hall/CRC,1984.