|
Data mining is an important topic in the field of information processing , it combines a number of areas such as databases , artificial intelligence , machine learning , statistical theory and technology . Classification is one of the important functions of data mining , decision tree - based classification algorithm in data mining is very broad . Compared with other classification algorithms , decision trees have a relatively small amount of calculation , easy to extract explicit rules , you can display important decision attributes and classification accuracy high . In the practical application of the process , however , the existing decision tree algorithm also there are a lot of inadequacies , such as the calculation inefficient , spanning tree larger . Therefore, further improvement of the decision tree algorithm , to improve the performance of the decision tree , to make it more suitable for the application requirements of the data mining technology has important theoretical and practical significance . For these shortcomings , this paper carried out in-depth research , the rough set theory is introduced which decision tree classifier , to explore how to optimize the decision tree classification algorithm . This paper studies the work are as follows : First, the paper introduces data mining technology and theoretical foundation , and focus on the decision tree generated after pruning algorithm analysis and comparison . Secondly , from both attribute reduction and pruning of the decision tree algorithm is optimized based on Prune the attribute dependency attribute reduction algorithms ER and decision tree based on rough set theory after pruning algorithm . Finally, the optimized decision tree algorithm used in supplier evaluation system which , and the algorithm C4.5 algorithm were compared to verify the effectiveness of the algorithm .
|