Timeline for Clustering customer data stored in ElasticSearch
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 15, 2014 at 5:52 | vote | accept | Konstantin V. Salikhov | ||
| S May 15, 2014 at 5:49 | history | suggested | buruzaemon | CC BY-SA 3.0 |
Just tightening things up a bit to improve readability
|
| May 14, 2014 at 23:51 | review | Suggested edits | |||
| S May 15, 2014 at 5:49 | |||||
| May 14, 2014 at 22:40 | answer | added | Nick Peterson | timeline score: 6 | |
| May 14, 2014 at 16:15 | comment | added | Konstantin V. Salikhov | I've read about SVM and I think its more about classification of newly created data after manual training on existing dataset - not about clustering existing data and finding abnormally big clusters. Am I right? If I am then this method isn't what I want. | |
| May 14, 2014 at 9:31 | comment | added | yatul | It isn't clear how you will perform preparement steps of your data. But you should look at algorithms described at en.wikipedia.org/wiki/Anomaly_detection . If I were you, I've checked SVM method first | |
| May 14, 2014 at 9:11 | comment | added | Konstantin V. Salikhov | Interesting groups are any groups of size greater than some threshold an that are much bigger than other possible clusters. | |
| May 14, 2014 at 9:08 | comment | added | yatul | What do you mean under "interesting groups"? Do you have some predefined important feature list? | |
| May 14, 2014 at 9:07 | history | edited | Konstantin V. Salikhov | CC BY-SA 3.0 |
fix typo
|
| May 14, 2014 at 8:38 | history | asked | Konstantin V. Salikhov | CC BY-SA 3.0 |