Timeline for Do infrequent examples screw up classifiers? If so, when is it okay to remove the infrequent examples from the data?
Current License: CC BY-SA 3.0
2 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 15, 2011 at 16:44 | comment | added | paul | Thanks for the response. I meant it in the way you describe first, not as an outlier. These are data from an experiment I designed, and I think the reason for the infrequent examples is probably a flaw in the experimental design. So while I am very interested in the infrequent classes, it seems like there's not enough data on them because of the design, not because of some real fact about the phenomena I'm studying. I could be wrong and it could be like the fraud detection scenario, but I think I can make an argument for why that's not the case when I present the results. | |
| Jun 14, 2011 at 23:29 | history | answered | doug | CC BY-SA 3.0 |