摘要
本文方法通过学习诸多信源在一定时间段内的变化过程 ,挖掘出那些与结论相关的信源及与结论相关时间片段 .形成最终的决策树模式 .
The relative sources and the time period can be defined by learning all features changes in a period.The set of examples is groups of time series of source data.A decision tree is constructed by the information gain.After the coherence estimation,the same as time series,the irrelative sources are washed out and relative sources are selected to be applied for fusion.The method fit the large scale fusion of dynamic hidden.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2002年第2期292-294,共3页
Acta Electronica Sinica
基金
国家自然科学基金 (No .698730 0 7)
关键词
信息融合
示例学习
决策树
动态相关性挖掘
信源
data fusion
learning from examples
decision tree
mining
dynamic association