A K-Motifs Discovery Approach for Large Time-Series Data Analysis

摘要：

　　Motif discovery is a method for finding some previously unknown but frequently appearing patterns in time series.However,the high dimensionality and dynamic uncertainty of time series data lead to the main challenge for searching accuracy and effectiveness.In our paper,we propose a novel k-motifs discovery approach based on the Piecewise Linear Representation and the Skyline index,which is superior to traditional R-tree index.As the experimental results suggest,our approach is more accurate and effective than some other traditional methods.

关键词： Time series Data mining K-motifs Indexing and retrieval

作者: Yupeng Hu Cun Ji Ming Jing Xueqing Li

作者单位: School of Computer Science and Technology,Shandong University,Jinan,China

会议类型: 国际会议

会议名称: International Asia-Pacific Web Conference(第18届国际亚太互联网大会)

会议地点: 苏州

会议语种:英文

页码: 492-496

在线出版日期: 2016-09-23（万方平台首次上网日期，不代表论文的发表时间）