The Adjusted Rand Index is a measure of similarity between a clustering and some ground-truth that is adjusted for chance. It considers all pairs of samples that are assigned in the same or different clusters in the predicted and empirical clusterings.
$$ {\displaystyle ARI = {\frac {\left.\sum {ij}{\binom {n{ij}}{2}}-\left[\sum {i}{\binom {a{i}}{2}}\sum {j}{\binom {b{j}}{2}}\right]\right/{\binom {n}{2}}}{\left.{\frac {1}{2}}\left[\sum {i}{\binom {a{i}}{2}}+\sum {j}{\binom {b{j}}{2}}\right]-\left[\sum {i}{\binom {a{i}}{2}}\sum {j}{\binom {b{j}}{2}}\right]\right/{\binom {n}{2}}}}} $$
Estimator Compatibility: Clusterer
Score Range: -1 to 1
This metric does not have any parameters.
use Rubix\ML\CrossValidation\Metrics\RandIndex;
$metric = new RandIndex();