作者: Elke Achtert , Thomas Bernecker , Hans-Peter Kriegel , Erich Schubert , Arthur Zimek
DOI: 10.1007/978-3-642-02982-0_35
关键词: Pattern recognition 、 Mathematics 、 Visualization 、 Distance measures 、 Artificial intelligence 、 Dynamic time warping 、 Algorithm 、 Range (statistics) 、 Series (mathematics) 、 Order of integration 、 Time series 、 Edit distance
摘要: ELKI is a unified software framework, designed as tool suitable for evaluation of different algorithms on high dimensional real-valued feature-vectors. A special case feature-vectors are time series data where traditional distance measures like L p -distances can be applied. However, also broad range specialized like, e.g., dynamic time-warping, or generalized second order distances, shared-nearest-neighbor have been proposed. The new version 0.2 now extended to and offers selection these measures. It serve visualization- evaluation-tool the behavior data.