Hey, R-listers I am going to approximate arbitrary 1-D data density by mixture of Gaussian models. The problem is that given a set of data generated from an unknown density function, and want to use a Gaussian mixture density model to approximate it. Now how to determine the number of components, the means and variances for all the components? And can I plot them with plot of original empirical data density?
Thanks for your advice. Fred ______________________________________________ [EMAIL PROTECTED] mailing list http://www.stat.math.ethz.ch/mailman/listinfo/r-help