Skip to content

[source]

F-Beta#

A weighted harmonic mean of precision and recall, F-Beta is a both a versatile and balanced metric. The beta parameter controls the weight of precision in the combined score. As beta goes to infinity the score only considers recall, whereas when it goes to 0 it only considers precision. When beta is equal to 1, this metric is called an F1 score.

Fβ=(1+β2)precisionrecall(β2precision)+recall

Estimator Compatibility: Classifier, Anomaly Detector

Output Range: 0 to 1

Parameters#

# Name Default Type Description
1 beta 1.0 float The ratio of weight given to precision over recall.

Example#

use Rubix\ML\CrossValidation\Metrics\FBeta;

$metric = new FBeta(0.7);

Last update: 2021-01-26