水利信息化Issue(6)：34-41,54,9.DOI:10.19364/j.1674-9405.2025.06.006
基于模体的水文多元时间序列关联规则挖掘方法研究
Research on motif-based mining method for hydrological multivariate time series association rules
摘要
Abstract
Existing methods for mining association rules in hydrological multivariate time series generally face problems such as information loss due to symbolic representation,inability to capture temporal patterns,low efficiency,and a large number of redundant rules.To address these limitations,this study introduced the concept of time series motifs and proposed a method for mining method for hydrological multivariate time series association rules based on constraints and frequent pattern trees.The motif results were converted into multivariate pattern series,and efficient rule generation was achieved through temporal constraints.Experimental results indicated that the proposed motif-based mining method for hydrological multivariate time series association rules performed better than similar methods in terms of effectiveness,scalability,and robustness.It efficiently generated temporal association rules,highlighted the significance of temporal features in hydrological multivariate time series association rules mining.The research demonstrated strong predictive performance and practical value in hydrological data experiments.关键词
多元时间序列/模体挖掘/频繁模式/时态关联规则/水文数据Key words
multivariate time series/motif mining/frequent pattern/temporal association rules/hydrological data分类
信息技术与安全科学引用本文复制引用
SHI Yongqiang,REN Jianbo,MAO Jiaxiang,WANG Sheng,YANG Wen,FENG Jun..基于模体的水文多元时间序列关联规则挖掘方法研究[J].水利信息化,2025,(6):34-41,54,9.基金项目
国家重点研发计划项目(2024YFC3210800) （2024YFC3210800）
