Research Of Classification Algorithms Based On Decision Tree

Data mining means the process of incomplete, noisy, fuzzy and random data, extraction of implicit, in which people don't know in advance, but potentially useful information and knowledge. In this process, the classification of data in data mining is an important subject of study.This paper expounds the technical data mining and the classification of basic theory, mainly introduces how to use the decision tree to data mining method of classification, decision tree is described in detail the basic knowledge and related algorithms, and several typical algorithms of decision tree analysis and compared. The main research work as follows:1. Analysis of XML technology and the basis of the decision tree, a tree in the XML database mining in the analysis model, which can effectively solve the problem of different data interface.2. ID3 algorithm based on the attribute values tend to be more disadvantages, introducing the probability of influencing factors of ID3 algorithm corrected, make the decision tree to reduce the dependence of the attribute values, and through the use of students' information collection of two kinds of decision tree algorithm of comparison, achieved good effect.3. Using the modified decision tree algorithm, using C++ language, In meteorological technical guarantee information system in data mining, and to provide a rational decision-making departments, the scientific decision-making basis.