Research On The Classifying Algorithm Based On Decision Tree

Learning algorithm of decision tree bases on the instances, and focuses on deducing classification rules represented by decision tree from a group of out-of-order and out-of-rule samples. And the decision tree has significant theoretical and practical value in the research fields of artificial intelligence, such as machine learning, data mining and intelligent control.In this paper the algorithm of building decision tree is discussed and researched, and some significant conclusions are achieved.From the point of view of analyzing and comparison of the decision tree algorithms, the classical algorithms of ID3 and C4.5 are descripted, then the advantages and disadvantages of two algorithms are analyzed and compared. The algorithm of ID3 usually leans to select attributes which value much more, and can't handle continuous data, and is sensitivity to the noise. The algorithm of C4.5 is an optimized one based on the ID3 algorithm. It can process the attributes which have continuous value and empty value data. Whereas the C4.5 algorithm tends to select the attribute with minimum entropy value, not that contributes most to classification.As for the selection of test attribute, we analyses how to acquire the support degree of correct partition in the case of partitioning the sample data set with condition attributes. And the support degree is the capability of making the correct decisions with the condition attributes. Based on the idea aforementioned, the support degree of the decisions is defined as the ratio of two numbers of pairs, and one is the number of pairs of elements which are distinguishable using the condition attribute with respect to the decision attribute, the other is the number of pairs of elements needed to be...