Chinese Journal of Lasers, Volume. 48, Issue 3, 0311002(2021)

Recognition of Food-Borne Pathogenic Bacteria by Raman Spectroscopy Based on Random Forest Algorithm

Qi Wang1, Wandan Zeng1、*, Zhiping Xia2、*, Zhiping Li2, and Han Qu2
Author Affiliations
  • 1College of Computer Science and Information Engineering, Shanghai Institute of Technology, Shanghai 201418, China
  • 2Military Veterinary Institute, Changchun, Jilin 130062, China
  • show less
    Figures & Tables(12)
    Original Raman spectra
    Raman spectra after normalization
    Raman spectrum after Savitzky-Golay processing
    Pareto chart of principle components
    Frame of random forest algorithm
    Model accuracy change with n_estimators
    Model accuracy change with max_depth
    10-fold cross-validation diagram
    • Table 1. CICC numbers and names of eleven food-borne pathogenic bacteria

      View table

      Table 1. CICC numbers and names of eleven food-borne pathogenic bacteria

      NumberLatin name
      10869Yersinia enterocolitica
      10870Klebsiella pneumoniae
      21482Salmonella enterica subsp. enterica serovarInfantis
      21530Escherichia coli EHEC O157:H7
      21534Shigella flexneri
      21560Cronobacter sakazakii
      21600Staphylococcus aureus
      21617Vibrio parahaemolyticus
      22933Acinetobacter baumannii
      22956Salmonella enterica subsp. enterica serovarTyphimurium
      23794Vibrio cholerae
    • Table 2. Work process of decision tree algorithm

      View table

      Table 2. Work process of decision tree algorithm

      Decision tree algorithm
      Input: sample X, sample numbers N, feature counts M
      Output:Decision Tree model
      X→for bagging∥processing X with bagging cycles
      end for
      while extracting ntry(ntry=N)→Xtrain do
      Mmtry(mtryM)∥ random selection of mtry attributes
      mtry→the best node
      XXsamples// build samples using Bootstrap
      end while
      for (itree=0; 1<itreeNtree; itree++)
      ∥ node splitting by optimal attributes to generate decision trees
      end for
      end procedure
    • Table 3. Process of algorithm used for prediction of food-borne pathogenic bacteria

      View table

      Table 3. Process of algorithm used for prediction of food-borne pathogenic bacteria

      Food-borne pathogenic bacteria prediction algorithm
      Input: sample X, training set Xtrain, test set Xtest
      Output: K trees, prediction result r
      for all i = 1 to K do
      while jN do
      rowsample=rowsample+selectXtrain
      j++
      end while
      while stop condition not true do
      colsample-selectrowsample
      split_Attribute-min{Gini(colsample)}
      ∥ classification attributes are determined by the minimum Gini value
      tree-AddNodesplit_Attribute
      end while
      leaf_node←node
      end for
      for all i-1 to K do
      Ri=Ti_PredictDtest
      R=MostCommon(Ri)
      end for
      end procedure
    • Table 4. Accuracy of each model

      View table

      Table 4. Accuracy of each model

      ModelAccuracy /%
      PCA + KNN(K-nearest neighbors)88.19
      PCA + logistic regression88.25
      PCA + SVM(support vector machines)83.86
      PCA + decision tree82.63
      PCA + RF(our)91.36
    Tools

    Get Citation

    Copy Citation Text

    Qi Wang, Wandan Zeng, Zhiping Xia, Zhiping Li, Han Qu. Recognition of Food-Borne Pathogenic Bacteria by Raman Spectroscopy Based on Random Forest Algorithm[J]. Chinese Journal of Lasers, 2021, 48(3): 0311002

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: spectroscopy

    Received: Jul. 6, 2020

    Accepted: Sep. 14, 2020

    Published Online: Feb. 2, 2021

    The Author Email: Zeng Wandan (zengwd@sit.edu.cn), Xia Zhiping (zengwd@sit.edu.cn)

    DOI:10.3788/CJL202148.0311002

    Topics