Consensus clustering in complex networks

By Andrea Lancichinetti, Santo Fortunato. Scientific Reports, 2012

883 citations recorded by the source corpus.

The community structure of complex networks reveals both their organization and hidden relationships among their constituents. Most community detection methods currently available are not deterministic and their results typically depend on the specific random seeds, initial conditions and tie-break rules adopted for their execution. Consensus clustering is used in data analysis to generate stable results out of a set of partitions delivered by stochastic methods. Here we show that consensus clustering can be combined with any existing method in a self-consistent way, enhancing considerably both the stability and the accuracy of the resulting partitions. This framework is also particularly suitable to monitor the evolution of community structure in temporal networks. An application of consensus clustering to a large citation network of physics papers demonstrates its capability to keep track of the birth, death and diversification of topics.

Subfields: Artificial Intelligence, Cognitive Neuroscience, Statistical and Nonlinear Physics

View Paper: https://www.semanticscholar.org/paper/7999b1b449fe6c9e2dcec0d05b6433407ec9aa9c

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 Cognitive Neuroscience

In Statistical and Nonlinear Physics

Citations in the corpus

Community detection in networks: A user guide (2016, methodology): “) Source: Reprinted figure with permission from [75].”
Structure and function of complex brain networks (2013, background): “(F) Mutliple applications of the modularity optimization algorithm (here 1000) yielded a number of unique solutions (here 47) that are displayed in the form of the consensus matrix.(41) Different gray levels refer to the number of times each node pair was placed into the same module.”
Modular Brain Networks (2015, methodology): “Lancichinetti & Fortunato (2012) advocated exploring a range of thresholds, because smaller or larger values of τ may produce correspondingly larger or smaller consensus communities.”
The Human Thalamus Is an Integrative Hub for Functional Brain Networks (2016, background): “A consensus matrix (a value of 1 where the two ROIs are in the same module and a value of 0 elsewhere; (Lancichinetti and Fortunato, 2012) was formed based on this result.”
Benchmarking of participant‐level confound regression strategies for the control of motion artifact in studies of functional connectivity (2017, methodology): “Because this approach to community detection is degenerate, a consensus partition was obtained over 100 Louvain optimizations (Lancichinetti and Fortunato, 2012).”
Cognitive Network Neuroscience (2015, methodology): “Consensus clustering methods can be utilized to uncover a modular decomposition that represents a group of subjects or time points (Lancichinetti & Fortunato, 2012), and such methods can be made more statistically rigorouswith the use of appropriate null models for comparison (Bassett, Porter, et…”
On the permanence of vertices in network communities (2014, methodology): “A huge volume of work has been devoted to ﬁnding communities in large networks, including diverse methods such as modularity optimization [3, 6], spectral graph-partitioning [23], random-walk [24], information-theoretic [25, 26], consensus clustering [17] and many others (see [8] for the review).”
Morphometric Similarity Networks Detect Microscale Cortical Organisation And Predict Inter-Individual Cognitive Variation (2017, positive): “We next resolved the modular community structure of MSNs by partitioning the sample mean MSN from coarse- to fine-grained scales defined by the resolution parameter (γ) of a consensus (1000 runs) Louvain modularity algorithm (Lancichinetti and Fortunato, 2012) (Figure 2C; Figure S2C).”
Network attributes for segregation and integration in the human brain. (2013, background): “New methodological developments in network science, for example involving link-based communities [37] and consensus clustering [38], may soon offer a new perspective on brain network organization.”
Multi-scale brain networks (2016, neutral): “(2012) and variants of the “association-recluster” framework (using a clustering algorithm to find consensus communities in a co-occurrence or association matrix that stores the frequency with which nodes co-occur in a community over an ensemble of partitions) (Lancichinetti and Fortunato, 2012; Bas”
Open Community Challenge Reveals Molecular Network Modules with Key Roles in Diseases (2018, methodology): “However, instead of selecting individual modules according to a quality score, ensemble methods merge alternative clusterings to obtain potentially more robust consensus predictions (Lancichinetti and Fortunato, 2012).”
Intrinsic and task-evoked network architectures of the human brain. (2014, neutral): “To identify a representative partition, several consensus algorithms exist (Bassett et al., 2013a; Doron et al., 2012; Lancichinetti and Fortunato, 2012).”
Graph theory methods: applications in brain networks (2018, background): “30 application of such a consensus approach can also reveal additional facets of community organization, including the degree to which individual nodes are affiliated with their host community.”