Constructing Cost Sensitive Decision Trees Based on Multi-Objective Optimization

We propose a multi-objective optimization based on the cost sensitive decision tree building method. The misclassification cost, test cost, waiting time cost and information gain rate as four optimization goals by using the method of linear weighting are adopted to transfer the multiobjective optimization problem into a single objective optimization problem, as the splitting attribute selection criterion; and then we put forward the specific strategy of building the minimum cost decision tree and a hybrid testing decision tree method; finally, use our algorithm and two other algorithms in two real datasets to build and test the decision trees. The experimental results show that our method of decision tree features less cost, more efficient and stronger generalization ability. The method is especially useful in terms of medical diagnostic.