gaussianMle

MLE (maximum likelihood estimator) for Gaussian PDF

Contents

Syntax

Description

gPrm = gaussianMle(data, type) returns the optimum parameters for Gaussian PDF via MLE

• data: data matrix where each column corresponds to a vector of observation
• type: type of covariance matrix
  • 'full' for full covariance matrix
  • 'diagonal' for diagonal covariance matrix
  • 'single' for a covariance matrix of a constant times an identity matrix
• gPrm: Parameters for Gaussian PDF
  • gPrm.mu: MLE of the mean
  • gPrm.sigma: MLE of the covariance matrix

Example

1D example

dataNum=1000;
x = randn(dataNum, 1);
showPlot=1;
gPrm=gaussianMle(x, [], showPlot);

2D example

ds=dcData(7);
figure; gPrm=gaussianMle(ds.input, 'full', showPlot);
figure; gPrm=gaussianMle(ds.input, 'diagonal', showPlot);
figure; gPrm=gaussianMle(ds.input, 'single', showPlot);

Published with MATLAB® R2018a