23_emgroup: Other Initializations

In EMCluster: EM Algorithm for Model-Based Clustering of Finite Mixture Gaussian Distribution

Other Initializations

Description

Two more functions with different initialization method.

Usage

starts.via.svd(x, nclass = 1, method = c("em", "kmeans"),
               EMC = .EMC)
emgroup(x, nclass = 1, EMC = .EMC)

Arguments

x	the data matrix, dimension n\times p.
nclass	the desired number of clusters, K.
method	method with the svd initializations.
EMC	the control for the EM iterations.

Details

The starts.via.svd utilizes SVD to initial parameters, and the emgroup runs the EM algorithm starting from the initial.

Value

The starts.via.svd returns an object with class svd, and the emgroup returns and object emobj with class emret.

Author(s)

Wei-Chen Chen [e-mail] and Ranjan Maitra.

References

https://www.stat.iastate.edu/people/ranjan-maitra

See Also

init.EM, .EMControl.

Examples

library(EMCluster, quietly = TRUE)
set.seed(1234)
x1 <- da1$da

emobj <- emgroup(x1, nclass = 10)
summary(emobj)

ret.0 <- starts.via.svd(x1, nclass = 10, method = "kmeans")
summary(ret.0)

EMCluster documentation built on Sept. 11, 2024, 5:27 p.m.