Expand description
Functions for scoring a set of predictions, i.e. evaluating how close predictions and truth are. All functions in this module obey the convention that higher is better.
Functionsยง
- accuracy
- Returns the fraction of outputs which match their target.
- f1
- Returns the f1 score for 2 class classification.
- neg_
mean_ squared_ error - Returns the additive inverse of the mean-squared-error of the outputs. So higher is better, and the returned value is always negative.
- precision
- Returns the precision score for 2 class classification.
- recall
- Returns the recall score for 2 class classification.
- row_
accuracy - Returns the fraction of outputs rows which match their target.