Clustering by mixture decomposition

• The patterns are drawn from a population with a known number of clusters.
• The probability density function for class is where is a vector of unknown parameters for .
• If is the a priori probability of class , the mixture density can be given by Summation of