Cluster Ensembles --- A Knowledge Reuse Framework for Combining Multiple Partitions

By Alexander Strehl, Joydeep Ghosh. 2002

5,108 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Learning to cluster in order to Transfer across domains and tasks (2017, methodology): “Looking at the top-left heat map in ﬁgure 8, which has D = 1 and 10 clusters, it can be observed that the NMI score is very robust to low similar pair recall, even lower than 0.5.”
Community detection in networks: A user guide (2016, methodology): “19 Strehl and Ghosh introduced an earlier definition of NMI, where the mutual information is divided by the geometric average of the entropies [81].”
User Profile Preserving Social Network Embedding (2017, methodology): “On Ego-Facebook, UPP-SNE significantly outperforms TADW, yielding 14% and 11% performance gains on accuracy and NMI.”
Similarity network fusion for aggregating data types on a genomic scale (2014, methodology): “Compatibility of data sources can be checked via normalized mutual information (NMI).”
Interpersonal Synchrony: A Survey of Evaluation Methods across Disciplines (2012, methodology): “[126] A. Strehl and J. Ghosh, “Cluster ensembles a knowledge reuse framework for combining multiple partitions,” in Journal on Machine Learning Research, 2002, pp. 583–617.”
A Convergent Functional Architecture of the Insula Emerges Across Imaging Modalities (2012, unknown): “Consensus clustering is a popular approach within the statistics and machine learning literatures (e.g., Monti et al., 2003; Steinley, 2008; Strehl and Ghosh, 2002), and has been successfully applied to iFC data in several previous studies (Bellec et al.”
A Comprehensive Survey of Clustering Algorithms (2015, methodology): “There are mainly 4 kinds of methods to get the set of initial clustering results as follows: Table 16 Consensus functions Name Typical algorithm and application Based on co-association matrix [85] Based on graph partition CSPA, HGPA and MCLA [86] Based on relabeling and voting VM [88] Based on the h”
Ensemble based systems in decision making (2006, background): “Some of the more promising ones include using ensemble systems in non-stationary environments [19], in clustering applications [94]–[96], and countless work on theoretical analyses of such systems, such as on incremental learning, improving diversity, bias-variance analysis, etc.”
Multi-class Classification without Multi-class Labels (2019, methodology): “We also include the normalized mutual information (NMI) (Strehl & Ghosh, 2002) metric.”
Fast hierarchical Bayesian analysis of population structure (2018, methodology): “A binary similarity matrix is then produced as described in Strehl and Ghosh (33) where for each pair of sequences (i,j), the (i,j)th entry in the matrix is 1 if the ith and jth sequences are clustered together and 0 otherwise.”
Time-series clustering - A decade review (2015, unknown): “in various studies [26,237,238].”
Stable Community Cores in Complex Networks (2012, methodology): “There are also similar methods in ensemble clustering like [19] but here we study networks, i.”
A Mixture Model for Clustering Ensembles (2004, methodology): “Strehl and Ghosh [30] have considered three different consensus functions for ensemble clustering: one is similar to the evidence accumulation approach [11, 12], and the other two methods are based on a hypergraph representation.”
Combining multiple clusterings using evidence accumulation (2005, methodology): “consensus clustering, as proposed in [56], under the assump-”
Stochastic Multiple Choice Learning for Training Diverse Deep Ensembles (2016, background): “Other approaches train or combine ensemble members under a joint loss [19, 26].”
Parallel Spectral Clustering in Distributed Systems (2011, methodology): “In practice, we made use of the following formulation to estimate the NMI score [38]:”