Comparing clusterings---an information based distance

By Marina Meila. 2007

1,734 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics

View Paper: https://www.semanticscholar.org/paper/c33ef60880d83232064f010ff3a9d0a5f20fe0eb

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

In Computational Theory and Mathematics

Citations in the corpus

Maximum Margin Temporal Clustering (2012, background): “These criteria are referred as pair-counting measures (Meila, 2007).”
A Convergent Functional Architecture of the Insula Emerges Across Imaging Modalities (2012, positive): “For each value of K, we compared the cluster solutions generated for each modality using two metrics: Percent Agreement and Variation of Information (VI; Meila, 2007).”
Community detection in networks: A user guide (2016, methodology): “Adjusted indices have also the disadvantage of nonlocality (Meilă, 2005): the similarity between partitions differing only in one region of the network depends on how the remainder of the network is subdivided.”
Random Walks, Markov Processes and the Multiscale Modular Organization of Complex Networks (2008, methodology): “A popular way to compare two partitions P 1 and P 2 is the normalized variation of information [79] where H ( 1”
On testing for spatial correspondence between maps of human brain structure and function (2018, positive): “Although statistics such as the adjusted Rand index (Hubert and Arabie, 1985) and mutual information (Meil a, 2007) can quantify the extent of overlap, there is no standard way to interpret the significance of these statistics that takes into consideration the spatial dependence of brain maps.”
Superhuman Accuracy on the SNEMI3D Connectomics Challenge (2017, methodology): “For the post-challenge analyses on AC3, we adopted the variation of information (VI), an information theoretic metric, to measure segmentation quality [23, 34].”
Unsupervised Incremental Online Learning and Prediction of Musical Audio Signals (2015, methodology): “Although this assumption will not always be true in our evaluation, we will use ARI, since it is a more established measure than alternative ones, such as the Fowlkes-Mallows index, the Mirkin metric, the Jaccard index [23], or entropy-based measures [32, 35], e.”
Weight-conserving characterization of complex functional brain networks (2011, methodology): “We quantified the similarity of individual partitions with the variation of information, a popular informationtheoretic measure of distance in partition space (Meila, 2007).”
Normalized Mutual Information to evaluate overlapping community finding algorithms (2011, background): “Meila [3] deﬁned a measure based on mutual information for comparing disjoint clusterings.”
Data-Driven Tree Transforms and Metrics (2017, methodology): “We compare the clustering of the samples relative to the refined subtype labels [42] using three measures: the Rand index (RI) [47], the adjusted Rand index (ARI) [48], and the variation of information (VI) [49].”
Functional Brain Networks Develop from a “Local to Distributed” Organization (2009, methodology): “The robustness of the community assignments was also tested using a different information theoretic procedure implemented by Meila, [95], which utilizes the measure ‘variation of information (VOI)’ (see Figure S7 and also [96]).”