基于矩阵轮廓的时间序列Shapelet发现算法
Time series Shapelet discovery algorithm based on matrix profile
陶琴 1杨骏 2王兵 3敬思远2
作者信息
- 1. 西南石油大学计算机科学学院,四川成都 610000;乐山师范学院特殊教育语言智能四川省哲学社会科学重点实验室,四川 乐山 614000
- 2. 乐山师范学院特殊教育语言智能四川省哲学社会科学重点实验室,四川 乐山 614000;乐山师范学院电子信息与人工智能学院,四川 乐山 614000
- 3. 西南石油大学计算机科学学院,四川成都 610000
- 折叠
摘要
当前时间序列Shapelet发现算法普遍采用穷举法,需要计算所有时间序列子序列的信息增益,效率较低.针对此问题,提出一种基于矩阵轮廓的Shapelet发现算法.选出最具代表性的时间序列对,计算其轮廓矩阵和差异向量,找到一簇关键区域;对找到的关键区域进行剪枝;在关键区域上搜索Shapelet并计算其信息增益,提升算法效率.在15个UCR数据集上,通过时间序列二分类实验对所提Shapelet发现算法进行验证.实验结果表明,所提算法结合Shapelet转换后具有较强分类能力,计算效率明显优于现有Shapelet发现算法.
Abstract
Current algorithms of time series Shapelet discovery,which are widely based on enumeration,require to figure out the information gain of time series subsequence and suffer from low efficiency.To handle this problem,an algorithm of Shapelet discovery based on matrix profile was proposed.The pairs of time series that were representative were selected out,their matrix profile and difference vector were calculated,and a set of critical areas was found.The critical areas were pruned.The Shapelet in the critical areas was searched and the information gain was calculated,improving the efficiency of the algorithm.Experiments of binary classification of time series were performed to evaluate the proposed Shapelet discovery algorithm.The results show that the proposed algorithm combined with Shapelet transformation has strong classification ability,and its efficiency is obviously superior to the existing algorithm of Shapelet discovery.
关键词
时间序列/二分类/模式发现/矩阵轮廓/关键区域/差异向量/信息增益Key words
time series/binary classification/pattern discovery/matrix profile/critical area/difference vector/information gain引用本文复制引用
基金项目
四川省科技计划重点研发基金项目(2021YFS0019)
厅市共建智能终端四川省重点实验室开放基金项目(SCITLAB-1002)
出版年
2024
NETL
NSTL
万方数据