Studying The Contrast Between Artificial Neural Networks And Decision Tree In Data Classification

With the maturity of database technology and the popularization of data application using in human society, data amount that mankind accumulate is increasing in incredibly rapid speed. Men are in face of the endless ocean of information and the huge wave of the data is billowing and welling up to people.On the other hand, the situation that contrasted sharply with the present is " We are drowning in information, but starving for knowledge." Part of the reason is due to lack the high-speed, high-efficient method making use of information technology to get useful information and knowledge. People are at a total loss what to do facing the huge amount of data.In order to solve the contradiction between the data enriching and poor knowledge, the researchers have developed Knowledge Discovery in Database (KDD) and its key technology—Data Mining (DM).Data classification is one of the most important methods in data mining. Have the numerous and disorderly data, people should classify these data first, then analyses and study the less data sets respectively, and then making decision and prediction.Based on the backgrounds mentioned above, we study two kinds of popular and important data classification methods—decision tree and artificial neural networks. We introduce their definitions, construction method and models. According to the practical conditions, both of the two methods are carefully and completely studied in contrast from different aspects and different levels.The conclusion that this paper gets finally is, under the specific environment and specified conditions, as a kind of new technology, the artificial neural networks method has distinctive predominance in solving the data classification problems.