Timeline for Logistic regression metric
Current License: CC BY-SA 4.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 14, 2022 at 20:14 | comment | added | Marko Lalovic | Because it does not differentiate between the number of correctly classified examples of different classes. Therefore, it can lead to wrong conclusions. Specifically, say we have an unbalanced data set with very few positive examples. On this data set, the majority classifier, which assumes class 0 for all cases, can achieve very high accuracy, although it does not return any relevant results. On the other hand this is reflected in zero recall (sensitivity). | |
| Mar 14, 2022 at 15:03 | comment | added | Dave | What's misleading about accuracy? | |
| Mar 14, 2022 at 12:40 | history | answered | Marko Lalovic | CC BY-SA 4.0 |