Comparing Clusterings by the Variation of Information

By Marina Meila. Annual Conference Computational Learning Theory, 2003

756 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/3696da952ee2f16793094455fdff394023b80e5c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Comparing clusterings: an axiomatic view (2005, background): “dM (1̂, C U K) = 1 − 1 K With respect to convex additivity, it is easy to prove (see also (Meilă, 2003)) that dD is convexely additive, while the Mirkin metric satisfies”
CUDA-Based Parallelization of Power Iteration Clustering for Large Datasets (2017, methodology): “The chosen cluster quality indexes were RI (Rand Index) [22], Jaccard [23] and Meil [24].”
Proxy Non-Discrimination in Data-Driven Systems (2017, background): “This quantity can be estimated from a sample in time O(|D | + k |Z |), where k is the number of elements in the range of p1 [46].”
Clustering Stability: An Overview (2010, background): “…deﬁned on the same data points, then it is straightforward to compute a distance score between these clusterings based on any of the well-known clustering distances such as the Rand index, Jaccard index, Hamming distance, minimal matching distance, Variation of Information distance (Meila, 2003).”
An extensive comparative study of cluster validity indices (2013, methodology): “In order to avoid the possible bias introduced by the selection of a particular partition similarity measure, we replicated all the experiments using three partition similarity measures: Adjusted Rand [31], Jaccard [41] and Variation of Information [42].”
Rethinking LDA: Why Priors Matter (2009, methodology): “In order to measure similarity between two sets of topic assignments Z and Z ′ forW , we can compute the distance between these partitions using variation of information (VI) [11, 6] (see suppl.”
Fixed-Rate Compressed Floating-Point Arrays (2014, background): “the variation of information [34] (a metric) between segmentations X and Y , H(X,Y ) is the joint entropy of corresponding segmentation labels, and I(X;Y ) is their mutual information.”
Metrics Reloaded: Recommendations for image analysis validation (2022, neutral): “A.G. triggered changes in the framework by responding to public questionnaire, was an active member of the calibration expert group, reviewed the manuscript, and participated in surveys.”
Evolutionary clustering (2006, background): “There are many examples of measures for comparing clusterings — see [13] for a discussion of the complexities that arise even in the case of flat clusterings.”
Beyond pixels: A comprehensive survey from bottom-up to semantic image segmentation and cosegmentation (2015, methodology): “Another commonly used region based criterion is the Intersection-over-Union, by checking the overlap between the S seg and S hum : • Variation of Information (VI): This metric [158] measures the distance between two segmentations S seg and S hum using average conditional entropy.”
Cut, Glue, & Cut: A Fast, Approximate Solver for Multicut Partitioning (2014, methodology): “The VI column reports the Variation of Information [20].”
Cluster ensembles (2011, background): “VI is another information theoretic measure proposed for cluster validation, 21 and deﬁned as It turns out that VI is a metric.”