Timeline for Silhouette Score not robust when clustering time series with tslearn
Current License: CC BY-SA 4.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 7, 2022 at 17:56 | history | edited | kjetil b halvorsen♦ | CC BY-SA 4.0 |
edited tags
|
| Apr 19, 2022 at 6:36 | comment | added | bk_ | @Dilip Upadhyay different random states should yield the same result...if that is not the case then the results are not robust, thats what I mean by this question | |
| Jan 20, 2022 at 11:15 | comment | added | Dilip Upadhyay | Different results could be due "random_state" paramter not set in. Try passing as like -- km = TimeSeriesKMeans(n_clusters=n, metric="dtw", random_state=0) | |
| Feb 21, 2021 at 18:08 | comment | added | user312088 | if the series is dynamical, you could use correlation dimension. another option is to look at minimum entropy in the histogram | |
| Jul 25, 2020 at 20:22 | comment | added | ttnphns | I'm not testing your code but giving a very general comment. Any internal clustering criterion should better analyzed graphically for sharp "elbows" rather than looking only on it extremum value (read "Comparing different k: priority of sharpness over extremum" here. Also, regular ("robust") results can be ever expected only when (1) there are relatively clear-cut clusters in the population, and (2) the (representative) sample is sufficiently large. There may be other nuances, too. | |
| Jul 24, 2020 at 14:08 | comment | added | bk_ | opened an issue on github: github.com/tslearn-team/tslearn/issues/278 | |
| Jul 24, 2020 at 7:45 | history | edited | bk_ | CC BY-SA 4.0 |
edited body
|
| Jul 23, 2020 at 14:53 | history | edited | bk_ | CC BY-SA 4.0 |
added 9 characters in body
|
| Jul 23, 2020 at 14:33 | history | asked | bk_ | CC BY-SA 4.0 |