Research On Decision Tree Algorithm And Application Based On Data Mining

Decision tree is a method of data classification mining based on information theory. First, we will build a decision tree by the medium of known training data. Then, capitalized on the decision tree which had been created, we predict with this data. The process of building the decision tree is considered as the process of data regularity inference, so we can believe that decision tree accomplished the visualization of data regularity and the output result of decision tree can be easily appreciated. The way of decision tree often have been used, because the decision tree method have a lot of merit, for example the accurate degree of this method is high and the result can be appreciated and efficiency is also higher. So far, there are many algorithms have been given and we can gain a prediction model by analyzed known catalog information. According to the type and character of mined information, we need use different catalog project and used most suitable and optimization algorithm to fill data mining needs. This is the emphasis of this essay and chiefly researching content.In this paper, we list the application of ID3 algorithm and knowledge optimization choosing strategy. Attribution recursion of information gain optimization method applied to the researching of the entrance examination for college student determine first time, and according to actual requirement we have already test it on this field. This is a new field application and also is the first creation by now.First time create the Attribution recursion of information gain optimization algorithm in combination with ID3 algorithm knowledge optimization choosing strategy. Experiment show that this algorithm is the advance of ID3 algorithm and gain binary tree. So we integrate two algorithms with developing respective advantage and gain a well performance optimization algorithm.Through comparison decision tree between new algorithm which attribution recursion optimization method in combination with knowledge prevail choosing strategy and classical ID3 algorithm, we can discovered that the decision tree based on the new amendment algorithm is more brevity and more accessible to the ideal decision tree.