Timeline for What classification algorithm should one use after seeing that t-SNE separates classes well?
Current License: CC BY-SA 3.0
13 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 11, 2017 at 20:17 | history | bounty ended | Tomek Tarczynski | ||
| Apr 11, 2017 at 20:17 | vote | accept | Tomek Tarczynski | ||
| Apr 10, 2017 at 15:01 | comment | added | Tomek Tarczynski | (+1): Thanks for the answer, I was looking exactly for this. I will accept the answer within two days. | |
| Apr 10, 2017 at 14:53 | history | edited | Zhubarb | CC BY-SA 3.0 |
deleted 9 characters in body
|
| Apr 10, 2017 at 14:29 | history | edited | Zhubarb | CC BY-SA 3.0 |
deleted 2 characters in body
|
| Apr 10, 2017 at 12:55 | history | edited | Zhubarb | CC BY-SA 3.0 |
deleted 1 character in body
|
| Apr 10, 2017 at 12:52 | comment | added | Zhubarb | Please see edit, instance-based learning methods would be compatible with any clues you may get from t-SNE plots. | |
| Apr 10, 2017 at 12:50 | history | edited | Zhubarb | CC BY-SA 3.0 |
added 678 characters in body
|
| Apr 10, 2017 at 12:44 | comment | added | Tomek Tarczynski | Berkmeister: Everything what you wrote make sense. My problem is quite easy to describe: I want to make classification model as good as possible. How can I use it to improve the quality of model? My best guess was that t-SNE can 'suggest' which method might be relevant to given problem. | |
| Apr 10, 2017 at 12:29 | history | edited | amoeba | CC BY-SA 3.0 |
added 40 characters in body
|
| Apr 10, 2017 at 12:12 | history | edited | Zhubarb | CC BY-SA 3.0 |
added 168 characters in body
|
| Apr 10, 2017 at 11:56 | history | edited | Zhubarb | CC BY-SA 3.0 |
added 90 characters in body
|
| Apr 10, 2017 at 11:50 | history | answered | Zhubarb | CC BY-SA 3.0 |