Gaussian Mixture Models Based on Principal Components and Applications

Table 5

Forensic glass fragments data: parameter estimates and BIC of the bivariate GMM (first technique).

Values	Model parameters
	First component					Second component
	Parameter estimation	0.24311	1.58580	−0.62982	1.97795	3.97284	0.75689	−0.50935	0.20229	0.37507	0.20665
Log-likelihood = −531.5947, BIC = 1111.483; Number of iterations: 27