ELKI in Time: ELKI 0.2 for the Performance Evaluation of Distance Measures for Time Series

ELKI is a unied software framework, designed as a tool suitable for evaluation of dierent algorithms on high dimensional realvalued feature-vectors. A special case of high dimensional real-valued feature-vectors are time series data where traditional distance measures like Lp-distances can be applied. However, also a broad range of specialized distance measures like, e.g., dynamic time-warping, or generalized distance measures like second order distances, e.g., shared-nearestneighbor distances, have been proposed. The new version ELKI 0.2 now is extended to time series data and oers a selection of these distance measures. It can serve as a visualization- and evaluation-tool for the behavior of dierent distance measures on time series data.

• Type: Conference
• Year: 2009
• Where: SSD
• Authors: Elke Achtert, Thomas Bernecker, Hans-Peter Kriegel, Erich Schubert, Arthur Zimek
• Keywords: Database, Distance Measures, SSD 2009, Time Series Data, Traditional Distance Measures