英语论文网

g classification, nearest neighbor classification method , in which the most widely used decision tree algorithm . Decision tree is similar to the flow chart of a directed graph , it is the expression of knowledge . Decision tree algorithm in the training set by learning to get a tree , find some valuable , potential information . Decision tree algorithm has the following advantages : First, the decision tree algorithm complexity is small, you can generate a decision tree with great speed ; Second , decision tree algorithm anti-noise ability, many decision tree algorithm can handle noisy data , missing value data, etc. ; Third, the decision tree algorithm to extract the rules simple, easy to understand , is conducive to classify prediction ; Fourth, the decision tree algorithm is scalable , can handle both small data set , it can handle vast amounts of data , to meet the needs of the real class prediction . Because of this, the decision tree algorithm in the clinical diagnosis of the disease , equity investment , product quality evaluation , the company has a very good control of operations and marketing decision-making role , based on decision tree classification algorithm has a high practical value.

Previous studies on the decision tree algorithm have been many achievements , different decision tree algorithm has its own advantages and disadvantages , have their own field of application . Therefore, according to results of previous studies to learn and master the core algorithm tree -related knowledge, to master a variety of common and Decision Tree pruning process , and summed up and compared the characteristics of each is necessary, the actual classification forecast has an important role in the application . In this paper, the actual demand for automotive quality evaluation , decision tree classifier through programming , quality assessment and classification of automotive forecasting , and the shortcomings of the original algorithm corresponding improvement methods , thereby reducing the scale of the decision tree to improve the prediction accuracy of classification . Theoretical research and practical application of data mining classification allows us to forecast a deeper understanding of the field .
本论文由英语论文网提供整理，提供论文代写，英语论文代写，代写论文，代写英语论文，代写留学生论文，代写英文论文，留学生论文代写相关核心关键词搜索。