Acronym: TWEDType: Whole SeriesYear: 2009Publication: PAMI

Description: Marteau propose the TWE distance, an elastic distance metric that includes characteristics from both LCSS and DTW. It allows warping in the time axis and combines the edit distance with L$_p$-norms. The warping is controlled by a stiffness parameter, $\nu$. Stiffness enforces a multiplicative penalty on the distance between matched points in a manner similar to WDTW. A penalty value $\lambda$ is applied when sequences do not match (Algorithm 1).
Source Code: Time Warp Edit Distance Code
Published Results:Recreated Results:

Published
Dataset:Result:
Adiac0.376
Beef0.533
CBF0.007
Coffee0.214
FaceAll0.189
FaceFour0.024
FiftyWords0.187
Fish0.051
GunPoint0.013
Lightning20.213
Lightning70.247
OliveOil0.167
OSULeaf0.248
SwedishLeaf0.102
SyntheticControl0.023
Trace0.05
TwoPatterns0.001
Wafer0.004
Yoga0.13

This algorithm doesn't have any recreated results.

Algorithm: