AN OPTIMIZATION FOR CLASSIFICATION MAXIMUM-LIKELIHOOD CRITERION

A clustering criterion introduced by Symons (1981), which is called Cl assification Maximum Likelihood (CML) criterion in this paper, is desi gned to consider the cluster size and the covariance structure of samp les. The CML criterion is optimized by the 'Moving method' suggested b y Duda and Hart (1973, p. 226). When the Moving method is applied to t he CML criterion with an arbitrary initial cluster, it often yields de generate clusters. To avoid such degenerate cases, we propose two stag es of clustering. In the first stage, we roughly partition samples wit h respect to 'the covariance structure component' in the CML criterion . The resulting partition is then further clustered with the full CML criterion.