Scoring

O-Cluster generates a Bayesian probability model for scoring new data based on discovered clusters.

The generated probability model is a mixture model where the mixture components are represented by a product of independent normal distributions for numerical attributes and multinomial distributions for categorical attributes.