3260 papers • 126 benchmarks • 313 datasets
This task has no description! Would you like to contribute one?
(Image credit: Papersgraph)
These leaderboards are used to track progress in time-series-averaging-14
No benchmarks available.
Use these libraries to find time-series-averaging-14 models and implementations
No subtasks available.
This work introduces trainable time warping (TTW), whose complexity is linear in both the number and the length of time- series, and compares TTW and GTW on S5 UCR datasets in time-series averaging and classification.
A new divergence is proposed, dubbed soft-DTW divergence, which is non-negative and minimized when the time series are equal, under conditions on the ground cost and also proposes a new "sharp" variant by further removing entropic bias.
Adding a benchmark result helps the community track progress.