摘要
为实现对不同鞋底材料的快速、准确鉴别,采用高灵敏度X射线荧光光谱对88个不同品牌、不同种类的鞋底材料进行测定。根据测试结果,初步把样品分为4类。结合化学计量学,对数据进行分析,首先用Ward法进行凝聚式层次聚类,然后建立决策树随机森林分类预测模型,数据表明随机森林的训练集正确率为100%,测试集的正确率为89%。进一步根据所含元素可以对鞋底材料进行准确识别。高灵敏度X射线荧光光谱法结合化学计量学可以对鞋底材料进行快速有效的检验,为未知鞋底材料溯源提供了一种新的思路。
In order to identify different sole materials rapidly and accurately,this research tests 88 sole materials with different brands and different types by high-sensitivity X-ray fluorescence spectroscopy.Based on the test results,the samples are preliminarily divided into 4 categories.Combined with chemoichiometrics,this research firstly conducts condensed hierarchical clustering with Ward method,and then establishes the decision tree random forest classification prediction model.Data shows that the accuracy rate of the training set of random forest is 100%,and the test set 89%.The sole material can be accurately identified based on the contained elements.High-sensitivity X-ray fluorescence spectrometry combined with chemometrics can be used to quickly and effectively examine shoe sole materials,which provides a new idea to identify the unknown sole materials.
作者
王会荣
姜红
宋彩芳
刘姝君
郑先云
李桂兰
韩玮
WANG Hui-rong;JIANG Hong;SONG Cai-fang;LIU Shu-jun;ZHENG Xian-yun;LI Gui-lan;HAN Weis(Department of Criminal Science and Technology,Shanxi Police College,Taiyuan 030401,China;Wanzijian testing technology(Beijing)Co,Ltd.Beijing 100141,China;School of Data,Shanxi Normal University,Taiyuan 030031,China;Institute of Environmental Science,Shanxi University,Taiyuan 030006,China;Beijing Anke Huisheng Technology Co.Ltd.,Beijing 100123,China)
出处
《化学研究与应用》
北大核心
2025年第3期655-661,共7页
Chemical Research and Application
基金
山西省科技厅项目(20210302123472)资助。
关键词
鞋底材料
高灵敏度X射线荧光光谱法
层次聚类
决策树
随机森林
sole material
high sensitivity X-ray fluorescence spectroscopy
hierarchical clustering
decision tree
random forest
