Timeline for Unsupervised outlier detection in 2D space
Current License: CC BY-SA 3.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 8, 2016 at 12:09 | comment | added | StepTNT | Updated the OP as I'm doing more tests, sorry if I keep on adding stuff but I need some guidance while exploring all the possibilities. | |
| Nov 4, 2016 at 21:31 | comment | added | Erich Schubert | You certainly can use a lower epsilon to see more in the plot; and you can try larger minpts. If you have many duplicates, a larger minpts may be necessary to prevent everything from being connected. | |
| Nov 3, 2016 at 17:12 | comment | added | StepTNT | Is there any chance that I can have the original code or the one that you wrote for Elki? I'd like to try it anyways. By the way I just added the results of OPTICSXi to the OP so that you can see why it doesn't look good to me. | |
| Nov 3, 2016 at 16:31 | comment | added | Erich Schubert | We have had an implementation of TURN*, and I also got the original code. But it would not work well for us, it is for pixel data. We never got it work well enough to be worth including. If I remember correctly, the method works on a discrete ("pixels") data model; "left side" neighbor is literally that, the next pixel to the left. Parameter-free clustering is an illusion. For TURN* the essential parameter is the discretization into pixels. | |
| Nov 3, 2016 at 13:44 | comment | added | StepTNT | I know that being able to choose parameters is good, but my assignment requires something that can run without inserting parameters. For example, the image in the OP contains more than 41k points while another dataset has a size of just 2500, so using the same minPts can be quite risky. Since you're the main programmer behind ELKI (and thank you for that software!), do you have any plan on adding TURN*? Because it looks really promising for what I need to do. | |
| Nov 3, 2016 at 12:04 | comment | added | Erich Schubert | You need to choose these to suit your problem. It's good to be able to do so. Clustering is an exploratory technique, you want parameters to explore. See e.g. this answer. But depending on your problem, parameters may transfer from one data set to another similar data set. So maybe you can use the same minpts (e.g. 10) for all of your data sets; epsilon is an upper bound to get a nicer plot, you do not need to set it. Xi is relative, so 5% may work for all of them, too. | |
| Nov 3, 2016 at 10:34 | comment | added | StepTNT | Doesn't this solution makes everything supervised? I mean, both epsilon and minPts need to be manually chosen somehow and I need to run this on different datasets, so I can't use any "magic" value. | |
| Nov 3, 2016 at 9:42 | history | edited | Erich Schubert | CC BY-SA 3.0 |
added 61 characters in body
|
| Nov 3, 2016 at 9:32 | history | answered | Erich Schubert | CC BY-SA 3.0 |