New answers tagged

0 votes

Why does precision decrease with inceasing threshold?

The jagged and falling-off nature of the precision-threshold graph in logistic regression can be attributed to several factors. One key factor is the nature of the data being used for the prediction. ...
RegressIt's user avatar
  • 316
1 vote

ROC curve manual calculation vs. pROC package R

If the pROC function yields a ROC curve with an area under the curve less than $1/2,$ pROC will flip the labels and recalculate ...
Dave's user avatar
  • 3,688
2 votes

Multinomial Logistic Regression sensitive to choice of Encoding

Most ML methods assume that you properly preprocessed the data, which includes mean removal and scaling. It is so, because the underlying optimization algorthms are indeed sensitive to scaling. So you ...
gabalz's user avatar
  • 156

Top 50 recent answers are included