Elastic similarity and distance measures for multivariate time series

Abstract This paper contributes multivariate versions of seven commonly used elastic similarity and distance measures for time series data analytics. Elastic can compensate misalignments in the axis data. We adapt two existing strategies a version well-known Dynamic Time Warping (DTW), namely, Independent Dependent DTW, to these measures. While be applied various analysis tasks, we demonstrate their utility on classification using nearest neighbor classifier. On 23 datasets, that each but one achieves highest accuracy relative others at least dataset, supporting value developing suite also there are datasets which either dependent all more accurate than independent counterparts or vice versa. In addition, construct neighbor-based ensemble show it is competitive other state-of-the-art single-strategy classifiers.