Timeline for why SVM is preferred over logistic regression?
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 6, 2017 at 1:07 | history | tweeted | twitter.com/StackStats/status/905235980705288192 | ||
| Sep 5, 2017 at 9:03 | answer | added | ksha | timeline score: 1 | |
| Sep 4, 2017 at 20:33 | comment | added | Divyanshu daiya | thanks for help man ,Mathew Drury,i thought same way, but wasn't able to assure myself | |
| Sep 4, 2017 at 20:31 | comment | added | Divyanshu daiya | sorry,can't get source,read in some forum, can't remember. | |
| Sep 4, 2017 at 20:29 | comment | added | Matthew Drury | Logistic regression predicts probabilities, not class assignments. So all you will find in that case is that the predicted probabilities are all small, which aligns well with the data. There is no reason to believe svm would do better, what you read is mistaken and overgeneralizing. | |
| Sep 4, 2017 at 20:27 | comment | added | Divyanshu daiya | suppose in data set of 1000 only 20 are positive and rest are negative ,so even without training logistic regression like for example even if we all time give answer as negative on new data , though we might be wrong but our success rate is high as their are very few positives in our data set,so even if our trained logistic regression gives false positives success rate is gonna be high(but by F1 scores we can test our algo),i read somewhere that SVM would perform better on such data set,but why? | |
| Sep 4, 2017 at 20:13 | comment | added | Matthew Drury | What exactly is "skewed data"? | |
| Sep 4, 2017 at 20:09 | comment | added | Firebug | Source for the claim? Neither is preferred, they excel in different things. Now, logistic regression really do penalize over all the support, while SVMs only penalize on/across the margin, so perhaps this might explain that claim. | |
| Sep 4, 2017 at 20:07 | review | First posts | |||
| Sep 4, 2017 at 22:44 | |||||
| Sep 4, 2017 at 20:07 | history | asked | Divyanshu daiya | CC BY-SA 3.0 |