For classification of unseen cases system LERS uses a modified bucket brigade algorithm . The decision to which class a case belongs is made on the basis of three parameters: specificity, strength, and support. They are defined as follows: Specificity is the total number of conditions of the rule, strength is the total number of cases correctly classified by the rule during training, and support is the sum of all products of specificity and strength for all matching rules from the class. For any case, the class with the largest value of support determines the case membership.
The first LERS classification system was implemented, among others, by Andi Yan. Independently, the classification system was re - implemented by Chien - Pei B. Wang, who also tested different possible classification strategies. Yet another research on classification strategies, taking into account different ways of using certain and possible rules, was performed by Xihong Zou.
Return to the Main Page