hparetomixt.init: Provides initial values for the parameters of a mixture of...

In condmixt: Conditional Density Estimation with Neural Network Conditional Mixtures

Description

Initial values for the parameters of a mixture of hybrid Paretos are provided by applying the following steps : 1) clustering the sample into as many clusters as there are mixture components 2) estimating the hybrid Pareto parameters for each component on the data from each cluster with the moment-like estimators, see hpareto.mme

Usage

hparetomixt.init(m, x, iter.max = 20, nstart = 10)

Arguments

m	number of mixture components
x	data sample from which the initial parameters are computed
iter.max	maximum number of iteration for kmeans clustering, default is 20, see kmeans
nstart	number of random cluster centers chosen (default is 10), see kmeans

Value

a matrix of dimension 4 x m which stores the 4 parameters (pi, xi, mu, sigma) of each of the m components.

Author(s)

Julie Carreau

References

Carreau, J. and Bengio, Y. (2009), A Hybrid Pareto Model for Asymmetric Fat-tailed Data: the Univariate Case, 12, Extremes

See Also

kmeans, hpareto.mme

Examples

r <- rfrechet(500,loc=5,scale=5,shape=5)
m <- 2
param.init <- hparetomixt.init(m,r)

condmixt documentation built on July 1, 2020, 6:04 p.m.