Mixed dissimilarity measure for piecewise linear approximation based time series applications

作者：

Highlights：

• We developed a novel dissimilarity measure for piecewise linear approximation.

• It considers both mean and slope of segments.

• It capable to provide as good results without DTW as alternate measures with DTW.

• Empirically tests reveal that considering multiple features lowers the chance of pathological warpings of DTW.

摘要

•We developed a novel dissimilarity measure for piecewise linear approximation.•It considers both mean and slope of segments.•It capable to provide as good results without DTW as alternate measures with DTW.•Empirically tests reveal that considering multiple features lowers the chance of pathological warpings of DTW.

论文关键词：Dynamic time warping,Local distance,Piecewise linear approximation,Similarity,Constraining

论文评审过程：Available online 10 June 2015, Version of Record 24 June 2015.

论文官网地址：https://doi.org/10.1016/j.eswa.2015.06.003