决策树方法在提高电脱盐效果中的应用  

Application of decision tree in enhancing effect of electric desalination

作　　者：郭超[1] 

机构地区：[1]南京工业大学信息科学与工程学院

出　　处：《计算机工程与设计》2010年第10期2372-2375,共4页Computer Engineering and Design

摘　　要：针对石化生产过程中原油脱盐效率降低的现状,首次提出了运用决策树方法来提高电脱盐的效果。介绍了C4.5决策树算法的思想及实现流程,并详细阐述了信息增益的概念及计算方法,在此基础上对电脱盐数据进行了预处理,建立了决策树模型,然后利用数据挖掘的分类功能提取出分类规则,找到了影响电脱盐效果的主要因素并提出了提高电脱盐效果的方案。实验结果表明了该方法的有效性。Aiming at the reducing of the efficiency of Crude oil-desalting in the process of petrochemical production,the use of decision tree to enhance the effect of electric desalting is presented at the first time.Firstly,the idea and process of C4.5 arithmetic one of base decision tree arithmetic are introduced and the concept and calculation method of information gain are provided in detail.Then,the data of electric desalting is pretreated and the model of decision tree is established.At last the classification rules are extracted.It finds the main factors of effecting on electric desalting and puts forward the program of improving the removal rate with the help of the classification of data mining.This method is proved effective by practice.

关 键 词：电脱盐 数据挖掘 决策树 C4.5 信息增益 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]