1<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/CrossValidation/Metrics/Informedness.php">[source]</a></span>
2
3# Informedness
4Informedness a multiclass generalization of Youden's J Statistic and can be interpreted as the probability that an estimator will make an informed prediction. Its value ranges from -1 through 1 and has a value of 0 when the test yields no useful information.
5
6$$
7{\displaystyle Informedness = {\frac {\text{TP}}{{\text{TP}}+{\text{FN}}}}+{\frac {\text{TP}}{{\text{TN}}+{\text{FP}}}}-1}
8$$
9
10**Estimator Compatibility:** Classifier, Anomaly Detector
11
12**Output Range:** -1 to 1
13
14## Parameters
15This metric does not have any parameters.
16
17## Example
18```php
19use Rubix\ML\CrossValidation\Metrics\Informedness;
20
21$metric = new Informedness();
22```
23
24## References
25[^1]: W. J. Youden. (1950). Index for Rating Diagnostic Tests.