A new approach to classification

Clustering is a common and important issue, and finite mixture models based on the normal distribution are frequently used to address the problem. In this article, we consider a classification model and build a mixture model around it. A good assessment of the allocation of observations and number of clusters is easily obtained from this approach