摘要
根据战场目标信息丰富、数据量大和需求多变等特点,提出了一种基于惩罚关联属性的改进决策树算法,以进行战场目标辅助研判。针对不同研判需求,通过对属性进行分组、取舍和修正,按照递归分析属性的重要程度,找寻目标最佳属性划分标准及属性与属性组的先后顺序,生成特定决策树,对目标进行分析判定,辅助指挥员进行决策。仿真试验表明,该算法在现代智能化战争中具有较好的应用价值。
Considering the characteristics of rich information,large amount of data and variable demand of battlefield targets,an improved decision tree algorithm based on penalty associated property is proposed to assist the battlefield target judgment.According to different judgment requirements,through grouping,selection and correction of properties,the importance of properties is recursively analyzed,the optimal attribute division standard and the sequence of properties and properties groups are inferred,a specific decision tree is generated to analyze the purpose and assist commanders to make decisions.The simulation results show that this method is effective and has great value in modern intelligent warfare.
作者
李亚钊
李文强
陈娜
肖海峰
LI Yazhao;LI Wenqiang;CHEN Na;XIAO Haifeng(The 28th research institute of china electronics technology group corporation,Nanjing 210007,China;Unit 31001 of PLA,Beijing 100094,China)
出处
《指挥信息系统与技术》
2020年第1期62-67,共6页
Command Information System and Technology
基金
全军共用信息系统装备预研课题资助项目。
关键词
决策树算法
关联属性
分组策略
战场目标
辅助研判
decision tree algorithm
associated property
grouping strategy
battlefield target
assistant judgment
