Abstract: Methods and systems for automatically identifying and selecting preferred classification and regression trees are disclosed. Embodiments of the disclosed invention may be used to identify a specific decision tree or group of preferred trees that are predictively consistent across train and test samples evaluated against at least one node-specific constraint imposed by the decision-maker, while also having high predictive performance accuracy. Specifically, for a tree to be identified as preferred by embodiments of the disclosed invention, the train and test samples when evaluated node-by-node must agree on at least one key measure of predictive consistency. In addition to this node-by-node criterion, the decision-maker may adjust selection constraints to permit selection of a tree having a small number of node-by-node consistency disagreements, but with high overall tree predictive performance accuracy.

Abstract: Methods and systems for automatically identifying and selecting preferred size classification and regression trees are disclosed. The invention is used to identify a specific decision tree or group of preferred size trees that are consistent across train and test samples in node-specific details that are often important to decision makers. Specifically, for a tree to be identified as preferred by this system, the train and test samples must both agree on key measures for every terminal node of the tree. In addition to this node-by-node criterion, an additional tree selection method may be imposed. Accordingly, the train and test samples rank order the nodes on a relevant measure in the same way. Both consistency criteria may be applied in a fuzzy manner in which agreement must be close but need not be exact.

Abstract: The present invention provides a method and system for automatically identifying and selecting preferred classification and regression trees. The invention is used to identify a specific decision tree or group of trees that are consistent across train and test samples in node-specific details that are often important to decision makers. Specifically, for a tree to be identified as preferred by this system, the train and test samples must both agree on key measures for every terminal node of the tree. In addition to this node-by-node criterion, an additional tree selection method may be imposed. Accordingly, the train and test samples rank order the nodes on a relevant measure in the same way. Both consistency criteria may be applied in a fuzzy manner in which agreement must be close but need not be exact.