摘要
本文根据宜居城市内涵建立宜居度评价指标体系,引入支持向量机(Support Vector Machine,SVM)理论,建立了C-classification类型和线性核函数结合的广西城市宜居分类预测SVM模型,并利用R语言实现分层随机抽样的技术,对训练集与测试集样本数据的随机性和差异性进行探究。模型结果显示:基于SVM理论的广西城市宜居度等级分类预测模型,可靠性强、预测准确率高,准确率高达92.86%。同时,城市宜居度等级分类预测模型程序化语言的实现,对保障工程后期的研究预测的可持续性具有参考意义。
Based on the connotation of livable cities,this paper establishes the index system of livability evaluation,introduces Support Vector Machine(SVM)theory,and establishes the SVM model of the classification and prediction of livability of Guangxi cities combined with C-class type and linear kernel function.Using R language to realize stratified random sampling,the randomness and difference of the sample data of training set and test set are explored.The model results show that the classification prediction model of urban livability based on SVM theory has high reliability,high prediction accuracy and high accuracy of 92.86%.At the same time,the realization of the programming language of the urban livability classification prediction model is of reference significance to the sustainability of the research prediction in the later period of the project.
作者
吴荣火
钟德炎
范明丽
林天潮
孙远通
黄萍蓉
WU Rong-huo;ZHONG De-yan;FAN Ming-li;LIN Tian-chao;SUN Yuan-tong;HUANG Ping-rong(College of Mathematics and Statistics,Yulin Normal University,Yulin,Guangxi 537000)
出处
《玉林师范学院学报》
2019年第5期26-34,共9页
Journal of Yulin Normal University
基金
广西高校中青年教师基础能力提升项目(2019KY0607)
玉林师范学院科研项目(2018YJKY29)
大学生创新创业训练计划项目(201910606116、201910606108)。
关键词
宜居度
分类
支持向量机
分层随机抽样
R语言
livability
classification
SVM
Stratified random sampling
R language
