摘要
为探讨随机森林方法对大米产地确证的有效性,找寻相邻区域的产地确证优化模型,利用原子吸收分光光度计对梅河及其相邻产区大米的矿物质元素(Cu、Zn、Fe、Mn、K、Ca、Na、Mg、Pb、Cd)含量进行了测定。采用R语言编写程序,通过strata函数实现训练集与测试集的数据划分,选用random Forest函数建立产地确证模型,由袋外误差估计进行模型优化。研究结果表明:建立的模型能够实现相邻区域大米样本的产地确证,对50个待测样本进行预测,准确率达96%。利用随机森林结合矿物质元素指纹技术可确证和追溯大米的产地来源,为相邻区域的地理标志大米及其他食品的产地确证研究提供方法参考。
In order to explore the validity of random forest algorithm to confirm the origin of rice,to find the optimal model of producing area in adjacent regions,the contents of Cu,Zn,Fe,Mn,K,Ca,Na,Mg,Pb and Cd in rice of Meihe and its adjacent regions were determined by atomic absorption spectrophotometer. The program is written in R language,the data of training set and test set are realized by strata function,the model of origin confirmation is selected by random forest function,and the model is optimized by out-of-bag estimate of error rate. The results show that the established model can confirm the origin of the rice samples in the adjacent regions,and the prediction accuracy of the 50 samples is 96%.The rice can be confirmed and traced by using the combination of random forest and mineral element fingerprinting technology. This method provides reference for the confirmation of the origin of rice and other foods in the adjacent regions.
引文
[1] Rossmann A. Determination of stable isotope rations in food analysis[J]. Food Reviews International,2001,17(3):347-381.
[2] Gremaud G,Quaile S,Piantini U,et al. Characterization of Swiss vineyards using isotope data in bornbination with trace element and classical parameters[J]. European Food Resources and Technology,2004,219:97-104.
[3] Schellenberg A,Chmielus S,Schlicht C,et al. Multielement stable isotope ratios(H,C,N,S)of honey from different European regions[J].Food Chemistry,2010,121(3):770-777.
[4] Cabaero A I,Recio J L,Rupérez M. Liquid chromatography coupled to isotope ratio mass spectrometry:a new perspective on honey adulteration detection[J]. Agric Food Chem,2006,54(26):9719-9727.
[5] Pilgrim T S,Watling R J,Grice K. Application of trace element and stable isotope signatures to determine the provenance of tea(Camellia sinensis)samples[J].Food Chem,2010,118(4):921-926.
[6]袁玉伟,胡桂仙,邵圣枝,等.茶叶产地溯源与鉴别检测技术研究进展[J].核农学报,2013(4):452-457.
[7]王朝辉,张亚婷,闵伟红,等.水稻品种对大米产地溯源判别正确性的影响[J].吉林农业大学学报,2017,39(1):113-119.
[8]于合龙,张恒维,刘杰,等.水稻质量安全可追溯系统[J].吉林农业大学学报,2018,40(1):120-126.
[9] Maione C,de Paula E S,Gallimberti M,et al. Comparative study of data mining techniques for the authentication of organic grape juice based on ICP-MS analysis[J]. Expert Systems with Applications,2016,49:60-73.
[10] Rohman A,Che Man Y B,Hashim P,et al. FTIR spectroscopy combined with chemometrics for analysis of lard adulteration in some vegetable oils Espectroscopia FTIR combinada con quimiometría para el análisis de adulteración con grasa de cerdo de aceites vegetales[J]. Cyta-Journal of Food,2011,9(2):96-101.
[11] Zheng L,Watson D G,Johnston B F,et al. A chemometric study of chromatograms of tea extracts by correlation optimization warping in conjunction with PCA,support vector machines and random forest data modeling[J]. Analytica Chimica Acta,2009,642(1):257-265.
[12] Ill-Min Chung,Jae-Kwang Kim,Jae-Keun Lee,et al. Discrimination of geographical origin of rice(Oryza sativa L.)by multielement analysis using inductively coupled plasma atomic emission spectroscopy and multivariate analysis[J]. Journal of Cereal Science,2015,65:252-259.
[13]赵海燕,郭波莉,张波,等.小麦产地矿物元素指纹溯源技术研究[J].中国农业科学,2010,43(18):3817-3823.
[14]夏立娅,申世刚,刘峥颢,等.基于近红外光谱和模式识别技术鉴别大米产地的研究[J].光谱学与光谱分析,2013,33(1):102-105.
[15]莫菲凡,范伟,周冀衡,等.基于近红外光谱和随机森林方法鉴别蜂蜜真伪[J].食品安全质量检测学报,2014(8):2430-2434.
[16]郝莉花,张平.近红外光谱技术在食品产地溯源中的应用研究进展[J].农产品加工,2016(24):54-57.
[17]陈秋生,张强,刘烨潼,等.矿质元素指纹技术在植源性特色农产品产地溯源中的应用研究进展[J].天津农业科学,2014(6):4-8.
[18]周志华.机器学习[M].北京:清华大学出版社,2016:179-180.
[19]董师师,黄哲学.随机森林理论浅析[J].集成技术,2013(1):1-7.
[20] Han Jiawei,Micheline Kamber,Pei Jian.数据挖掘概念与技术[M].范明,孟小峰,译.3版.北京:机械工业出版社,2012:236-239.