Rand index

The Rand index is a measure used to evaluate how well a clustering algorithm groups data points compared to a true or reference grouping. It considers all pairs of data points: if the algorithm groups a pair the same way as the reference, it counts as a correct match; if it differs, it's a mismatch. The index calculates the proportion of pairs that are either correctly clustered together or correctly separated. A Rand index of 1 means perfect agreement, while 0 indicates no agreement beyond random chance. It helps assess the accuracy of clustering results in a straightforward, interpretable way.