f1score

It measures the performance of a classification model in terms of the both precision and recall. F1 score can be interpreted (loosely) as the average of precision and recall, where the best value is 1 and the worst is 0.

Attention: Available only with Twin Activate commercial edition.

Syntax

Score = f1score(targets,predictions,average)

Inputs

targets
Actual label for each observation.
Type: double
Dimension: vector
predictions
Predicted value for each observation.
Type: double
Dimension: vector
average
Averaging strategy in case of multiclass classification. 'micro' (default), 'macro', 'none' are the possible values for average. If 'none' is chosen, per class metric is given as output.
Type: char
Dimension: string

Outputs

Score
f1score of the classifier.
Type: double
Dimension: scalar | struct (if 'none' is chosen)

Example

Usage of f1score

targets = [0 1 0 1];
predictions = [1 1 1 1];
score1 = f1score(targets, predictions);
score2 = f1score(targets, predictions, 'micro');
score3 = f1score(targets, predictions, 'macro');
score4 = f1score(targets, predictions, 'none');
printf('Micro: %f \n', score1);
printf('Micro: %f \n', score2);
printf('Macro: %f \n', score3);
printf('None : ');
disp(score4);
Micro: 0.666667 
Micro: 0.666667 
Macro: 0.666667 
None : 
0.666666667