Timeline for How to measure the similarity of k-means clustering using different datasets?
Current License: CC BY-SA 3.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 21, 2015 at 15:25 | comment | added | EngrStudent | K-means start with different initial dispositions, repeat. If they don't come to the same center, then membership is different. Also, they sometimes flap. Look at your convergence criteria there. A "learning rate", truncated slightly before convergence can tell you about the nature of variation in convergence is. The fundamental assumption of k-means is equal variance for all clusters. | |
| Jul 21, 2015 at 14:44 | answer | added | dcorney | timeline score: 1 | |
| Aug 15, 2014 at 9:28 | history | asked | Samo Jerom | CC BY-SA 3.0 |