«上一篇/Previous Article|本期目录/Table of Contents|下一篇/Next Article»

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

30

期数:

2020年07期

页码:

174-179

栏目:

应用开发研究

出版日期:

2020-07-10

文章信息/Info

Title:

Empirical Comparison of Decision Classifiers in Air Quality Data Analysis

文章编号:

1673-629X(2020)07-0174-06

作者:

付 悦¹ ; 夏小娜¹; 2; 3

1. 曲阜师范大学 统计学院,山东 曲阜 273165; 2. 曲阜师范大学 信息科学与工程学院,山东 日照 276826; 3. 曲阜师范大学 中国教育大数据研究院,山东 曲阜 273165

Author(s):

FU Yue¹ ; XIA Xiao-na¹; 2; 3

1. Institute of Statistics,Qufu Normal University,Qufu 273165,China; 2. School of Information Science and Engineering,Qufu Normal University,Rizhao 276826,China; 3. China Institute of Big Data for Education,Qufu Normal University,Qufu 273165,China

关键词:

判别分析; 决策树; 随机森林; 支持向量机; 决策分类器; 空气质量

Keywords:

discriminant analysis; decision tree; random forest; support vector machine; decision classifier; air quality

分类号:

TP39

DOI:

10. 3969 / j. issn. 1673-629X. 2020. 07. 037

摘要:

分别使用传统统计学方法与机器学习方法对北京市空气质量状况展开实证分析,研究“判别分析”“决策树”“支持向量机”和“随机森林”四种分类方法对于同一数据集的分类结果。 为了研究该问题,首先对传统分类方法构建的判别分析分类器与机器学习分类方法构建的三类分类器进行了理论介绍,随后以 4:1 的比例将北京市空气质量状况数据分为训练集与测试集,使用训练集构建分类器,对构建的分类器进行优化处理后,使用测试集进行分析及预测,根据预测结果对传统分类方法与机器学习分类方法进行详细实证对比和评价。 采用传统分类方法与机器学习分类方法分别构建数据分类器,并应用于空气质量数据的统计和分析中,可为其他周期以及其他地区的空气质量研究提供方法支持,也可为类似数据的有效分类提供技术指导。

Abstract:

The traditional statistical method and machine learning method are used to conduct empirical analysis on the air quality of Beijing respectively,and the classification results of discriminant analysis,decision tree,support vector machine and random forest for the same data set are studied. In order to study the problem,first of all,the discriminant analysis of traditional classification method to build a classifier with machine learning classification methods of three kinds of classifiers for the construction theory is introduced,and then in a ratio of 4:1 Beijing’s air quality data can be divided into training set and test set,with the training set to build a classifier,for the construction of the classifier after optimized,using test set for analysis and prediction. According to the predicted results,the traditional classification method and machine learn-ing classification methods are carried out empirical comparison and evaluation in detail. The traditional classification method and machine learning classification method are respectively used to construct data classifiers and apply them to the statistics and analysis of air quality data,which can provide methodological support for air quality research in other periods and other regions,and also provide technical guidance for effective classification of similar data.

常用功能

更新日期/Last Update: 2020-07-10