论文信息 - Similar subsequence retrieval from two time series data using homology search

Similar subsequence retrieval from two time series data using homology search

We propose a method for extracting the most similar subsequences from two time series data by quantizing them and performing a homology search. The homology searches, such as BLAST and SW, are string search algorithms. Therefore, time series data should be quantized. SAX and EIAD were applied as quantization methods, and their effectiveness was examined by experiment. According to the experiments, time series data sets were classified into four types of time series data set, and we discuss the characteristics of SAX and EIAD.

Tomoyuki Hiroyasu | Mitsunori Miki | Masato Yoshimi | Hisatake Yokouchi | Takuma Nishii

[1] Kunio Kashino,et al. A quick search algorithm for acoustic signals using histogram features—time‐series active search , 2001 .

[2] Oka Ryuichi,et al. Reference Interval-free Continuous Dynamic Programming for Spotting Speech Waves by Arbitrary Parts of a Reference Sequence Pattern. , 1996 .

[3] Eamonn J. Keogh,et al. Exact Discovery of Time Series Motifs , 2009, SDM.

[4] Li Wei,et al. Fast time series classification using numerosity reduction , 2006, ICML.

[5] 電子情報通信学会. The Transactions of the Institute of Electronics, Information and Communication Engineers , 1987 .

[6] Eamonn J. Keogh,et al. A symbolic representation of time series, with implications for streaming algorithms , 2003, DMKD '03.