针对现有学习方法对完全时间不对称数据的动态贝叶斯网络学习不具有实用性,提出一种借助传递变量进行完全时间不对称数据的动态贝叶斯网络结构学习方法.首先进行相邻时间片间的传递变量序列学习;然后,基于节点排序和局部打分-搜索,进行动态贝叶斯网络局部结构学习;最后通过时序扩展得到整个动态贝叶斯网络结构. The dynamic Bayesian network learning based on existing learning methods for asymmetric complete time data is not practical and a dynamic Bayesian network structure learning method based on full time asymmetric data transmission variables is proposed.Firstly, Then, based on node ordering and local scoring - search, dynamic Bayesian network local structure learning is carried out. Finally, the entire dynamic Bayesian network structure is obtained through time series expansion.