Pre-pruning Learning Algorithm For Decision Tree Based On Rough Set Theory

We do not yet know how to make computers learn nearly as well as people learn. However, not only in applications, algorithms, theory, but also in studies of biological systems, the rate of progress has increased significantly in current years. The research of intelligent knowledge-discovery plays a great role in these progresses.Decision tree learning is one of the widely used and practical methods for inductive inference. Decision tree represents the information of the original dataset by an explicit tree model. Its accuracy and comprehensibility depend on how concisely the learning algorithm can summarize this structure. The non-predictive parts of a decision tree should be eliminated or pruned. Pruning mechanisms require an instrument to detect the relationship between the decision tree and the training data. The measurement of the uncertainty of a decision table is an ideal tool for doing exactly that.The main methods to prune a decision tree are pre-pruning and post-pruning. Pre-pruning seem more direct, but it is very difficult to estimate precisely when to stop growing the tree. It always depends on the expert knowledge or prior knowledge. This reduces the intelligent level of the learning process, and limits its application. In this paper, we use the whole certainty of a decision table to control the process of the decision tree pre-pruning, and develop a knowledge discovery method which controlled by the data itself. The experiment results show that the method is efficient and feasible.