DECISION TREE INDUCTION BASED ON EFFICIENT TREE RESTRUCTURING

The ability to restructure a decision tree efficiently enables a varie ty of approaches to decision tree induction that would otherwise be pr ohibitively expensive. Two such approaches are described here, one bei ng incremental tree induction (ITI), and the other bei ng non-incremen tal tree induction using a measure of tree quality instead of test qua lity (DMTI). These approaches and several variants offer new computati onal and classifier characteristics that lend themselves to particular applications.