A window-based characterization method for biophysical time series


Tezin Türü: Doktora

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Enformatik Enstitüsü, Bilişim Sistemleri Anabilim Dalı, Türkiye

Tezin Onay Tarihi: 2017

Öğrenci: DENİZ KATIRCIOĞLU

Danışman: NAZİFE BAYKAL

Özet:

In thesis, we propose a robust similarity score-based time series characterization method, termed as Window-based Time series Characterization (WTC). Specifically, WTC generates domain-interpretable results and involves remarkably low computational complexity thereby rendering itself useful for densely sampled and populated time series datasets. In this study, we apply WTC to a proprietary action potential (AP) time series dataset on human cardiomyocytes and three precordial leads from a publicly available electrocardiogram (ECG) dataset. We, then, compare WTC with shapelet transform and fast shapelet transform (which constitutes an accelerated variant of the former), in terms of predictive accuracy and computational complexity. The results indicate that WTC achieves a slightly higher classification performance with significantly lower execution time when compared to its shapelet-based alternatives. With respect to its characterization capability, WTC has a potential to enable medical experts to explore definitive common trends in novel datasets.