The Accuracy Comparison of Time Series Classification in Vector Space between SAX and BOSS Methods: A Case Study of Electrocardiogram

Main Article Content

Napatsorn Kaewkla
Akarin Phaibulpanich


The electrocardiogram (ECG) is an important procedure used to diagnose heart disorders. However, the ECG may contain different types of noise due to various of factors, potentially resulting in diagnostic errors. This research compares Symbolic Aggregate Approximation in Vector Space (SAXVSM) and Bag of Symbolic Fourier Approximation Symbols in Vector Space (BOSSVS) methods for classifying ECG data with noise. To choose a suitable classification algorithm for ECG5000 dataset, which is available in the Physionet database. Four types of ECG noises were simulated and then added to the data as follow: 1) Electromyography (EMG) 2) Powerline Interference 3) Baseline Wander and 4) Composite at 25%, 50% and 100% levels for the performance comparison of the ECG classification between normal and abnormal heart rhythms with SAXVSM and BOSSVS. The results show that both algorithms have similar high performance for all 13 datasets: accuracy and F1 score are 97-99%, precision is 95-99%, and recall is 97-100%, but BOSSVS has a longer running time than SAXVSM.

Article Details

How to Cite
N. Kaewkla and A. Phaibulpanich, “The Accuracy Comparison of Time Series Classification in Vector Space between SAX and BOSS Methods: A Case Study of Electrocardiogram”, JIST, vol. 11, no. 2, pp. 49–61, Dec. 2021.
Research Article: Multidisciplinary (Detail in Scope of Journal)


Thailand Online Hospital. "การตรวจคลื่นไฟฟ้าหัวใจ (Electrocardiography)," 2021. [Online]. Available: [Accessed April 15, 2021].

J. Lin, E. Keogh, L. Wei, and S. Lonardi, "Experiencing SAX: a novel symbolic representation of time series," Data Mining and Knowledge Discovery, vol. 15, no. 2, pp. 107-144, doi: 10.1007/s10618-007-0064-z, 2007.

P. Senin and S. Malinchik, "Sax-vsm: Interpretable time series classification using sax and vector space model," in 2013 IEEE 13th international conference on data mining, IEEE, pp. 1175-1180, 2013.

P. Schäfer, "Bag-Of-SFA-Symbols in Vector Space (BOSS VS)," Zuse-Institut Berlin (ZIB), 2015.

P. Schäfer, "Scalable time series similarity search for data analytics," doi: 10.18452/17338, 2015.

P. Schäfer, "The BOSS is concerned with time series classification in the presence of noise," Data Mining and Knowledge Discovery, vol. 29, no. 6, pp. 1505-1530, doi: 10.1007/s10618-014-0377-7, 2014.

Y. Kaya and H. Pehlivan, "Classification of premature ventricular contraction in ECG," Int J Adv Comput Sci Appl, vol. 6, no. 7, pp. 34-40, doi: 10.14569/IJACSA.2015.060706, 2015.

K. M. Chang, "Arrhythmia ECG noise reduction by ensemble empirical mode decomposition," Sensors (Basel), vol. 10, no. 6, pp. 6063-80, doi: 10.3390/s100606063, 2010.

P. Senin. "Z-normalization of time series," 2021. [Online]. Available: [Accessed April 21, 2021].

Souspace. "รู้จัก Discrete Fourier Transform และ Fast Fourier Transform," 2021. [Online]. Available: [Accessed April 21, 2021].

V. V. Raghavan and S. M. Wong, "A critical analysis of vector space model for information retrieval," Journal of the American Society for information Science, vol. 37, no. 5, pp. 279-287, doi: 10.1002/(SICI)1097-4571(198609)37:5<279::AID-ASI1>3.0.CO;2-Q, 1986.

N. Srikong. "Cosine similarity," 2021. [Online]. Available: [Accessed April 21, 2021].