OverlapErrorByDensity: OverlapErrorByDensity

In DistributionOptimization: Distribution Optimization

Description

Similarity in GMM by Density

Usage

OverlapErrorByDensity(Means, SDs, Weights, Data = NULL, Kernels = NULL)

Arguments

Means	Means of the GMM Components
SDs	Standard Deviations of the GMM Components
Weights	Weights of the GMM Components
Data	Dataset that the GMM should be compared with
Kernels	if length(Kernels) = 1: amount of kernels if length(Kernels) > 1: kernels in dataspace at which the GMM Components will be compared with each other

Details

Calculates the similarity (overlap) between multiple modes in Gaussian Mixture Models. Kernels at equally distanced positions are used, if not explicitly given.

Value

List: OverlapError Error for estimating the maximal Overlap of AUC of PDFs of each pair of GMM Components Kernels Kernels that were used for comparing the GMM Components

Author(s)

Florian Lerch

Examples

Data = c(rnorm(50,1,2), rnorm(50,3,4))
V<-OverlapErrorByDensity(c(1,3), c(2,4), c(0.5,0.5), Data = Data)
AdaptGauss::PlotMixtures(Data, c(1,3), c(2,4), SingleGausses = TRUE)
print(V$OverlapError)

DistributionOptimization documentation built on Feb. 12, 2020, 5:57 p.m.