Lukasz A. Kurgan and Krzysztof J. Cios proposed a supervised CAIM (class-attribute interdependence maximization) discretization algorithm [5] that handles continuous and mixed mode attributes. The CAIM algorithm‟s goal is to find the minimum number of discrete intervals while minimizing the loss of class-attribute interdependency.
The algorithm uses class-attribute interdependency information as the criterion for the optimal discretization.
The Class- Attribute Interdependency Maximization (CAIM) criterion measures the dependency between the class variable C and the discretization variable D for attribute F, for a given quanta matrix. The CAIM criterion is defined as