clustering: Spectral clustering

In CAMCOS/camcos2017: CAMCOS Spring 2017: Efficient and Scalable Spectral Clustering for Document Clustering.

Description

Spectral clustering

Usage

clustering(Weights, method, k, t = NULL, sparse = TRUE,
  kmeans.method = "kmeans", m = NULL)

Arguments

Weights	Matrix of similarity weights. Can be dense or sparse.
method	NJW, Ncut, DiffusionMap.
k	Number of clusters.
t	Number of steps for the diffusion map.
sparse	Logical. Indicates whether the input for the weights argument is sparse.
kmeans.method	One of the following: "kmeans", "fuzzy", "poly.fuzzy"
m	Fuzziness parameter. Defaults to m=2 for fuzzy; m=.5 for poly.fuzzy.

Value

Integer vector. Cluster assignments for each row of original data.

See Also

similarity

CAMCOS/camcos2017 documentation built on May 6, 2019, 9:23 a.m.