基于DNA编码遗传算法构造广义决策树的研究  

Study on Constructing Generalized Decision Tree Based on DNA Coding Genetic Algorithm

作　　者：马冲[1] 

机构地区：[1]同济大学CAD研究中心,上海201804

出　　处：《计算机与数字工程》2009年第5期17-21,共5页Computer & Digital Engineering

摘　　要：决策树是归纳学习和数据挖掘的重要方法,主要用于分类和预测。文章引入了广义决策树的概念,实现了分类规则集和决策树结构的统一。同时,提出一种新颖的基于DNA编码遗传算法构造决策树的方法。先用C4.5算法对数据集进行分类得到初始规则集,再通过文章中算法优化规则集并由此构建决策树。实验证明了该方法有效地避免了传统决策树构建过程的缺点,且有较好的并行性。Decision tree is an important method of induction learning and data mining. It is mainly used in classification and predictive model. The introduction of generalized decision tree （GDT） realized the unification of classification rules and decision tree structure. Meanwhile, a new method that based on DNA coding genetic algorithm to construct deci- sion tree was proposed. It firstly classified dataset by C4.5 to get initial rule sets, then optimized the rule sets by using the algorithm to construct decision tree. It is proved that this method can avoid the weakness of traditional decision tree constructing process and has good parallelism.

关 键 词：DNA编码 遗传算法决策树广义决策树 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] TP18[自动化与计算机技术—计算机科学与技术]