Timeline for K-means clustering to separate temperature vertical profiles
Current License: CC BY-SA 4.0
2 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 12, 2020 at 16:52 | comment | added | enrymather | Actually, I don't have labels already. Depending on the weather situation, the percentage of stations belonging to either cluster could vary widely (from 50-50% to even 10-90%, I can imagine) and the two clusters could be centered anywhere in the region (W-E, S-N, NE-rest etc.). But I would like to keep the number of cluster fixed at 2 to avoid overcomplicating. So basically I'd like to use an algorithm which uses a certain threshold to decide whether the stations can be sorted into two separate clusters (i.e. vertical profiles) and if so, outputs these clusters. | |
| Nov 11, 2020 at 16:06 | history | answered | Marco Ramos | CC BY-SA 4.0 |