改进的C4.5算法及其在地质调查项目管理中的应用研究

发布时间：2018-03-14 02:36

本文选题：C4.5算法　切入点：决策树　出处：《兰州交通大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着各行业信息化的日益普及,所产生的数据量激增,但是却出现了“数据海量,知识贫乏”的局面。数据挖掘为解决这个问题提供了新的契机。数据挖掘是从大量的、复杂的、随机的数据之中提取有价值信息和知识的过程。在这个过程之中,对数据进行分类是数据挖掘研究领域比较重要的课题。目前用于分类的方法有很多种,比较经典的有决策树算法、遗传算法、关联规则、神经网络、粗糙集方法等。决策树算法因其分类准确率高、容易理解等优点被人们广泛应用。而C4.5算法是决策树算法中最为典型的算法之一,它不仅继承了CLS、ID3等算法的优点,而且对其部分缺陷进行了优化,使其在数据挖掘研究领域被众多研究者所广知。但是C4.5算法也存在某些缺陷。比如在决策树构造的过程中,需要对数据集进行多次的排序和顺序扫描,因而导致算法的低效以及过度分支等问题。鉴于此,本文针对C4.5算法存在的缺陷进行了优化,并以地质调查项目信息管理系统中已有的项目数据为基础,利用改进的C4.5决策树,挖掘出项目投入和项目最终成果之间的关系,为决策者提供了决策依据。本文的主要研究内容包括:(1)CLS算法、ID3算法和C4.5算法的性能分析和比较。通过比较经典算法的特点和优势,为改进C4.5算法提供理论基础;(2)C4.5算法的改进。针对C4.5算法的缺点,提出一种高效的改进C4.5算法;(3)利用改进C4.5算法,在地质调查项目管理可视化系统中进行数据挖掘应用。本文的主要创新点有:(1)提出了一种基于SQL的高效C4.5决策树算法的实现方案,并引入容错率的概念进一步提高了算法的执行效率;(2)利用优化后的C4.5算法对地质调查项目数据进行数据挖掘,并将数据挖掘的结果应用于地质调查项目管理可视化系统之中。本文的研究成果对于决策树算法的研究者和使用者有着重要的参考意义,为地质调查项目的管理者和决策者提供了决策的理论依据。
[Abstract]:With the increasing popularity of information technology in various industries, the amount of data generated has increased sharply, but there has been a situation of "massive data and poor knowledge". Data mining provides a new opportunity to solve this problem. The process of extracting valuable information and knowledge from complex, random data. In this process, data classification is an important topic in the field of data mining. The classical decision tree algorithm, genetic algorithm, association rule, neural network, rough set method and so on. The C4.5 algorithm is one of the most typical algorithms in the decision tree algorithm, which not only inherits the advantages of CLSN ID3 algorithm, but also optimizes some of its defects. It is widely known by many researchers in the field of data mining, but C4.5 algorithm also has some defects. For example, in the process of constructing decision tree, data sets need to be sorted and scanned several times. In view of this, this paper optimizes the C4.5 algorithm and makes use of the improved C4.5 decision tree based on the existing project data in the geological survey project information management system. The relationship between the project input and the final results of the project is excavated, which provides the decision basis for the decision makers. The main research contents of this paper include the performance analysis and comparison of the two algorithms, I. e. In order to provide a theoretical basis for improving C4.5 algorithm, an efficient improved C4.5 algorithm is proposed to improve C4.5 algorithm. Data mining is applied in the visualization system of geological survey project management. The main innovation of this paper is to put forward an implementation scheme of efficient C4.5 decision tree algorithm based on SQL. The concept of fault-tolerant rate is introduced to further improve the efficiency of the algorithm. The optimized C4.5 algorithm is used for data mining of geological survey project data. The results of data mining are applied to the geological survey project management visualization system. The research results in this paper have important reference significance for the researchers and users of decision tree algorithm. It provides a theoretical basis for decision-making for the managers and decision makers of geological survey projects.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P622;TP311.13

【相似文献】