Figure 5
From: Distribution Optimization: An evolutionary algorithm to separate Gaussian mixtures
Fit of a GMM with M = 3 modes to data set 3 (amplitudes of muscle potential evoked in healthy volunteers14). The distribution of the data is shown as probability density function (PDF) estimated by means of the Pareto density estimation (PDE23; black line) and overlaid on a histogram. The GMM fit is shown as a red line and the M = 3 single mixes are indicated as differently colored dashed lines (M#1, …, M#3). The Bayesian boundaries between the Gaussians are indicated as perpendicular magenta lines. At the right of the distributions, the respective QQ-plots are shown. Top: Original fit as published previously, obtained with an interactive EM based GMM adaptation14. Middle: Fit obtained with the automated “Distribution Optimization” algorithm. Bottom: Fit obtained using the EM algorithm without manual interaction. The figure has been created using the R software package (version 3.5.3 for Linux; http://CRAN.R-project.org/24) and the R libraries “AdaptGauss” (https://cran.r-project.org/package=AdaptGauss12) and “DistributionOptimization” (https://cran.r-project.org/package=DistributionOptimization).
