决策类划分的多变量决策树实例运算与优化分析  

Instance operation and optimization analysis of multivariable decision tree with decision class division

作　　者：黄俊南[1] HUANG Jun-nan(Quanzhou Vocational College of Economics and Business Infomlation Department,Fujian Quanzhou 362000,China)

机构地区：[1]泉州经贸职业技术学院信息系,福建泉州362000

出　　处：《齐齐哈尔大学学报（自然科学版）》2018年第5期1-6,共6页Journal of Qiqihar University(Natural Science Edition)

基　　金：福建省教育厅科技项目(JAT171059)

摘　　要：基于决策类划分的多变量决策树算法是一种新型的多变量算法。选取较复杂的实例构建训练集,并用新算法构造决策树,验证算法可行性和便捷性。从优化算法和比配分析两细节入手,有效地提升了算法的准确度,进一步降低了算法的时间复杂程度。The multivariable decision tree algorithm based on decision class division is a new multivariable algorithm. Select a more complex instance to build a training set and constructing a decision tree with a new algorithm, In this way, the feasibility and convenience of the algorithm are verified. From the two details of the optimization algorithm and comparative analysis, the accuracy of the algorithm is improved effectively, and the time complexity of the algorithm is further reduced.

关 键 词：决策树 多变量 训练集 样本优化 比配分析 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]