Timeline for Training multi-label classifier with low quality training set
Current License: CC BY-SA 3.0
17 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 2, 2017 at 13:16 | vote | accept | Lior Magen | ||
| Jan 2, 2017 at 12:44 | answer | added | SaiNageswar S | timeline score: 0 | |
| Jan 1, 2017 at 8:34 | 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. | |
| Dec 2, 2016 at 7:47 | 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. | |
| Nov 2, 2016 at 7:08 | answer | added | Emre | timeline score: 1 | |
| Nov 1, 2016 at 14:36 | 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. | |
| Oct 2, 2016 at 13:43 | 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 2, 2016 at 14:57 | comment | added | Jan van der Vegt | So for a document you know for a fact it belongs to A and B, for C you are not sure, but are there also topics you know for a fact it doesn't belong to? | |
| Sep 2, 2016 at 13:43 | 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. | |
| Aug 3, 2016 at 13:24 | answer | added | Hima Varsha | timeline score: 0 | |
| Aug 2, 2016 at 8:57 | comment | added | Lior Magen | Yes, just like in the title | |
| Aug 2, 2016 at 8:44 | comment | added | Hima Varsha | Are you referring to multilabel classification? | |
| Aug 2, 2016 at 8:04 | comment | added | Lior Magen | Did I made myself clear now? | |
| Aug 2, 2016 at 8:03 | comment | added | Lior Magen | If the document belongs to C I want it to be tagged as such but what about D, E and F? I don't know if the document belongs to D or doesn't belong to it. If I'll choose to tag the document to D (meaning that the output of class D will be 1) than it will be wrong, if I'll choose to tag the document as if it doesn't belong to D (meaning of class D will be 0) it will wrong as well. I'm not sure about class D so in such case I don't want the network to change its weights as a result of class D. | |
| Aug 1, 2016 at 7:09 | comment | added | Hima Varsha | I didn't get you. If the document belongs to C, why wouldn't you want to tag it with C during training? | |
| Jul 31, 2016 at 8:47 | review | First posts | |||
| Jul 31, 2016 at 9:11 | |||||
| Jul 31, 2016 at 8:46 | history | asked | Lior Magen | CC BY-SA 3.0 |