GroupSortFuse-package: The GroupSortFuse Package.

In tmanole/GroupSortFuse: Estimating the Number of Components in Finite Mixture Models

Description

Implementation of the Group-Sort-Fuse (GSF) method (Manole and Khalili 2019), which estimates the number of components in finite mixture models via penalized maximum likelihood estimation. The method is implemented for the following classes of mixture models,

• Location-Gaussian mixture models, with unknown or known covariance (normalLocOrder)

• Multinomial mixture models (multinomialOrder)

• Poisson mixture models (poissonOrder)

• Exponential distribution mixture models (exponentialOrder)

This package also provides coefficient path plotting functionality (plot.gsf). Tuning parameter selection can be performed using the Bayesian Information Criterion (bicTuning) or V-fold Cross Validation (cvTuning).

References

Manole, T., Khalili, A. 2019. "Estimating the Number of Components in Finite Mixture Models via the Group-Sort-Fuse Procedure".

tmanole/GroupSortFuse documentation built on Jan. 12, 2022, 10:37 p.m.