Quantifying morphology changes in time series data with skew
Author(s)Sung, Phil; Syed, Zeeshan; Guttag, John V.
MetadataShow full item record
This paper examines strategies to quantify differences in the morphology of time series while accounting for time skew in the observed data. We adapt four measures originally designed for signal shape comparison: Dynamic Time-Warping (DTW), Earth Mover's Distance (EMD), Frochet Distance (FD), and Hausdorff Distance (HD). These morphology difference metrics on time series are compared in discriminative power and noise resistance on ECG signals as well as on a synthetic dataset. We use data from our experiments to shed light on the relative strengths of the methods.
DepartmentMassachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory; Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
ICASSP (IEEE International Conference on Acoustics, Speech and Signal Processing)
Institute of Electrical and Electronics Engineers
Sung, P., Z. Syed, and J. Guttag. “Quantifying Morphology Changes in Time Series Data with Skew.” Acoustics, Speech and Signal Processing, 2009. ICASSP 2009. IEEE International Conference On. 2009. 477-480. Copyright © 2009, IEEE
Final published version
INSPEC Accession Number: 10700592