基于级联分类的复烤片烟产地预测方法研究
Study on prediction method of the geographical origin of flue-cured tobacco strips based on cascade classification
-
摘要: 为提高在有限样本条件下近红外预测复烤片烟产地的准确率,提出了一种基于级联分类的复烤片烟产地预测方法,该方法首先通过近红外光谱判断样本的香型属性并构建香型模型作为中间层,再使用LDA或PLS方法在单一香型框架下构建产地模型进行产地预测。以全国主要烟叶产区的复烤片烟为对象进行验证,结果表明:通过引入香型模型作为中间层,基于LDA的分类模型产地预测准确率由83.33%提升至94.44%;基于PLS的分类模型的准确率则由72.22%提升至86.11%。在有限样本数据和不引入新的模型参数的条件下,该方法有效降低了复烤片烟的产地误判比例。Abstract: In order to improve the accuracy of near-infrared prediction under limited sample conditions, a prediction method of the geographical origin of flue-cured tobacco strips based on cascade classification was proposed. In this method, the flavor attributes of samples were judged by near-infrared spectroscopy, and the flavor model was constructed as the middle layer, and then the origin model was constructed by LDA or PLS method under a single flavor framework to predict the geographical origin. The redried tobacco in the main producing area in China was taken as the objects in the experiment. The results showed that by introducing the flavor model as the middle layer, the accuracy of the classification model based on LDA was improved from 83.33% to 94.44%; the accuracy of PLS based classification model increased from 72.22% to 86.11%. Under the condition of limited sample data and without introducing new model parameters, this method could effectively reduce the proportion of misjudgment of the origin of redried tobacco.
-
-
[1]
李石头,廖付,何文苗,等.基于近红外光谱相似的烟叶替代与卷烟配方维护[J].烟草科技,2020,53(2):88-93.
-
[2]
杨建云,田孟玉,裴梓烨,等.基于近红外光谱技术的初烤烟叶致香成分含量特性快速定量评价[J].轻工学报,2018,33(5):60-68.
-
[3]
SHETTY N,GISLUM R,JENSEN A M D,et al.Development of NIR calibration models to assess year-to-year variation in total non-structural carbohydrates in grasses using PLSR[J].Chemometrics and Intelligent Laboratory Systems,2012,11(1):34-38.
-
[4]
彭彦昆,赵芳,李龙,等.利用近红外光谱与PCA-SVM识别热损伤番茄种子[J].农业工程学报,2018,34(5):159-165.
-
[5]
胡晓华,刘伟,刘长虹,等.基于太赫兹光谱和支持向量机快速鉴别咖啡豆产地[J].农业工程学报,2017,33(9):302-307.
-
[6]
王超,李朋成,杨凯,等.近红外光谱烟叶质量等级快速检测与等级特征分析[J].光谱学与光谱分析,2021,41(3):943-947.
-
[7]
施丰成,李东亮,冯广林,等.基于近红外光谱的PLS算法判别烤烟烟叶产地[J].烟草科技,2013,46(4):56-59.
-
[8]
孙文苹.基于近红外光谱的烟叶产地识别方法研究[D].青岛:中国海洋大学,2015.
-
[9]
束茹欣,蔡嘉月,杨征宇,等.应用近红外光谱投影模型法分析烟叶的产区与风格特征[J].光谱学与光谱分析,2014,34(10):2764-2768.
-
[10]
陈琦,汪爻,吴琼,等.基于元素及近红外分析的产地鉴别技术[J].现代食品,2020(3):165-171.
-
[11]
白雁,王星,张威,等.近红外光谱结合聚类分析法对不同区域连翘的鉴别研究[C]//中华中医药学会.中华中医药学会2009年药用植物化学与中药资源可持续发展学术研讨会.青海:[出版社不详],2009.
-
[12]
FU H Y,YIN Q B,XU L,et al.Challenges of large-class-number classification (LCNC):A novel ensemble strategy (ES) and its application to discriminating the geographical origins of 25 green teas[J].Chemometrics and Intelligent Laboratory Systems,2016,157(15),43-49.
-
[13]
国家烟草专卖局.烟草及烟草制品试样的制备和水分测定烘箱法:YC/T 31-1996[S].北京:中国标准出版社,1996.
-
[14]
BARNES R J,DHANOA M S,LISTER S J,et al.Standard normal variate transformation and de-trending of near-infrared diffuse reflectance spectra[J].Applied Spectroscopy,1989,43(5):772-777.
-
[15]
LI H F,JIANG T,ZHANG K S.et al.Efficient and robust feature extraction bymaximum margin criterion[J].IEEE Transactions on Neural Networks,2006,17(1):157-165.
-
[16]
CAI D,HE X F,HAN J W.Spectral regression:A unified approach for sparse subspace learning[C]//IEEE.Seventh IEEE International Conference on Date Miming (ICDM 2007),Omaha,[s:n],2007:9894636.
-
[17]
HAALAND D M,THOMAS E V.Partial least-squares methods for spectral analyses.2.Application to simulated and glass spectral data[J].Analytical Chemistry,1988,60(11):1202-1208.
-
[18]
王一丁,赵铭钦,付博,等.利用可见-近红外光谱鉴定不同香型风格烤烟的方法[J].中国烟草科学,2015,36(6):88-93.
-
[1]
计量
- PDF下载量: 20
- 文章访问数: 1546
- 引证文献数: 0