Timeline for Why does auto-encoder best suited for the job of anomaly detection?
Current License: CC BY-SA 4.0
17 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 5 at 16:04 | 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. | |
| May 26 at 6:07 | 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. | |
| Jan 25 at 9:07 | 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. | |
| Sep 25, 2023 at 14:07 | 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. | |
| May 17, 2023 at 19:09 | 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. | |
| Jan 17, 2023 at 1:02 | 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. | |
| Sep 17, 2022 at 15:03 | 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. | |
| May 19, 2022 at 14:29 | comment | added | EngrStudent | Not typical isn't the same as outlier. Domain expertise and root-causing is required here. Don't be making "black swans" for yourself. | |
| May 19, 2022 at 14:04 | 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. | |
| Jan 4, 2022 at 11:05 | 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. | |
| May 27, 2021 at 15:29 | answer | added | Single Malt | timeline score: 0 | |
| May 24, 2021 at 3:06 | review | First posts | |||
| May 24, 2021 at 15:14 | |||||
| May 17, 2021 at 16:05 | comment | added | Tylerr | The process of autoencoders naturally lends itself to outlier detection because if there is high reconstruction error then that means the model couldn't learn enough to fit that point. I.E. that point is an outlier. So, while other data points may be able to be represented by the bottleneck layer, that outlier point requires more information so it is different. But, as others mentioned, there are tons of other methods to do this and other applications of an autoencoder. | |
| May 17, 2021 at 15:01 | comment | added | Shaptak | Thanks. Whatever literature I have gone through regarding anomaly detection, mostly autoencoder is used. That's why I was wondering what properties of auto-encoder makes it more suitable for the purpose of anomaly detection ? | |
| May 17, 2021 at 14:58 | comment | added | mhdadk | Not necessarily. Check out Deep One-Class Classification. | |
| May 17, 2021 at 14:55 | comment | added | Tim | Is it? Who and where said it is? | |
| May 17, 2021 at 14:51 | history | asked | Shaptak | CC BY-SA 4.0 |