A Priority Heuristic Correlation Technique for Decision Tree Pruning

The main purpose of this study is to introduce a priority heuristic correlated information processing technique into decision support tree pruning problem by using behavior predicting and reasoning approach along with the probability correlation analysis framework. To verify its validity, evidence from some important tests is illustrated and discussed, and the results suggest that with this new technique, the size of decision tree can be structurally reduced without losing any predictive accuracy and make it better account for real-life decision-making problems.