分析测试学报2025，Vol.44Issue(6)：1190-1195,6.DOI:10.12452/j.fxcsxb.241201565
基于矿质元素含量结合机器学习的陈皮产地鉴别研究
Origin Identification of Citri Reticulatae Pericarpium Based on Miner-al Element Content Combined with Machine Learning
摘要
Abstract
The contents of mineral elements in 255 batches of Citri Reticulatae Pericarpium from Xin-hui and Guangxi were determined by inductively coupled plasma mass spectrometry.Orthogonal par-tial least squares discriminant analysis(OPLS-DA)was used to study the different elements in Citri Reticulatae Pericarpium from different producing areas.Four preprocessing methods,such as Z-score normalization,Min-Max normalization,mean normaliztion,and Max abs scaler,are used to establish a discriminant model by combining random forest(RF),decision tree(DT),support vector machine(SVM),and gradient boosting(GB)method.The results showed that among the 41 mineral elements,Na,Sn,Y,Ba,Er,Ho,Yb,Dy,Ni,Li,Gd,Tb,Sm,Nd,Rb were the main difference elements between Citri Reticulatae Pericarpium from Xinhui and Guangxi.Among the four machine learning models,the SVM model has the best prediction results.By SVM model,the accu-racy of the training group and test group under the three processing methods of Z-score normalization,Min-Max normalization,and mean normalization was the same,which was 100%and 96%,respec-tively,and the F1 value of 0.96.These result reflected the high accuracy of this method.Based on mineral element content combined with machine learning,this study established a high accuracy and reliability method for the origin identification of Citri Reticulatae Pericarpium,which provided techni-cal support for quality control of Citri Reticulatae Pericarpium and provided the basis for the origin traceability discrimination of traditional Chinese medicinal herbs.关键词
矿质元素/机器学习/陈皮/产地鉴别Key words
mineral elements/machine learning/Citri Reticulatae Pericarpium/origin identifica-tion分类
化学引用本文复制引用
周熙,刘倩宝,卢俏丽,张春华,康怀腾,刘畅,黄芳,吴惠勤,罗辉泰..基于矿质元素含量结合机器学习的陈皮产地鉴别研究[J].分析测试学报,2025,44(6):1190-1195,6.基金项目
广东省科学技术厅科技基础条件建设领域专项项目(2024B1212070001) （2024B1212070001）
广东省科学院发展专项资金项目(2022GDASZH-2022010110) （2022GDASZH-2022010110）
江门市科技特派员科研项目(2024760000280010017) （2024760000280010017）
