Efficient procedures for community detection in network studies, especially for sparse networks with not very obvious community structure. The algorithms impose penalties on the differences of the coordinates which represent the community labels of the nodes.
|Author||Yang Feng, Richard J. Samworth and Yi Yu|
|Date of publication||2013-12-15 17:52:32|
|Maintainer||Yi Yu <firstname.lastname@example.org>|
|License||GPL (>= 2.0)|
fcd: Fused community detection.
fcd.cluster: Clustering the estimators along the path.
fcd.criteria: The criterion values based along the path.
fcd.trans: The graph based penalty transformation matrix
generate: generate adjacency matrix of stochastic blockmodel,...
get.cluster: Final estimators of the community labels
isolate: Isolated nodes collection
laplacian: Laplacian matrix
mis.cluster: Mis-clustered nodes for balanced designed network.
spectral.clustering: Spectral clustering and its variant.