Timeline for How to evaluate the quality of a synthetic dataset?
Current License: CC BY-SA 3.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 9, 2019 at 19:03 | history | bumped | CommunityBot | This question has answers that may be good or bad; the system has marked it active so that they can be reviewed. | |
| Feb 4, 2019 at 5:02 | history | bumped | CommunityBot | This question has answers that may be good or bad; the system has marked it active so that they can be reviewed. | |
| Feb 19, 2018 at 7:08 | answer | added | Stephan Kolassa | timeline score: 1 | |
| Feb 18, 2018 at 20:23 | review | Close votes | |||
| Feb 19, 2018 at 13:59 | |||||
| Apr 25, 2017 at 15:33 | history | edited | Chill2Macht |
edited tags
|
|
| Nov 30, 2016 at 11:00 | comment | added | dimpol | There are different similarity measures that measure different things. Which one you need completely depends on what you want to do with it. It is like coming to us with an apple and saying: "I want things similar to this". It could be more apples, more fruit, more groceries, more red products, more products of that size/weight, more products of that price, etc. All are 'similar' in a way, which one you need is something only you know. | |
| Nov 30, 2016 at 10:54 | comment | added | Francesco | @dimpol I don't want an exact copy, I want to generate datasets that are at the same time similar to the original data but have their characterstics, for example, have peaks in different points or could show more/less noise. My objective is to find a way to say that synthetic dataset A is better/worse than synthetic dataset B. | |
| Nov 30, 2016 at 10:40 | comment | added | dimpol | This completely depends on what you want to use it for. Depending on the application, an exact copy of the original data or replacing all datapoints by the average for example might be good candidates or horrible options. | |
| Nov 30, 2016 at 9:57 | history | asked | Francesco | CC BY-SA 3.0 |