Laser & Optoelectronics Progress, Volume. 59, Issue 12, 1215011(2022)

Improved XGBoost Stray Current Prediction and Interpretable Model

Zhaoliang Meng1,2,3, Zetao Zhang1、*, Yuan Yang2, Guofeng Li3, Chongbo Tao3, and Yijiang Niu3
Author Affiliations
  • 1College of Electronic Information, Xi’an Polytechnic University, Xi’an 710600, Shaanxi , China
  • 2International Engineering College, Xi’an University of Technology, Xi’an 710048, Shaanxi , China
  • 3Power Electronics Division of CRRC Yongji Motor Co., Ltd., Xi’an 710000, Shaanxi , China
  • show less
    Figures & Tables(9)
    Error comparision under different characteristic number combinations
    Feature combination error contrast histograms. (a) MAE; (b) MSE; (c) RMSE; (d) R2
    Optimal characteristic iterative procession
    Comparison of different models for stray current prediction
    SHAP feature analysis sketch chart
    Dependence diagram of rail voltage SHAP analysis
    Dependence diagram of transition resistance SHAP analysis
    • Table 1. Error comparison between original model and optimized model

      View table

      Table 1. Error comparison between original model and optimized model

      ModelMSERMSEMAER2
      XGBoost8.4712.9103.4910.787
      OFS-XGBoost1.6841.2970.3340.923
    • Table 2. Prediction performance index of defferent models based on optimal feature combination method

      View table

      Table 2. Prediction performance index of defferent models based on optimal feature combination method

      ModelMSERMSEMAER2Prediction time /s
      SVR51.9087.2016.8880.6864.3
      BP7.3912.7732.3550.7752.4
      RF2.9451.7152.3830.8461.8
      OFS-XGBoost1.6841.2970.3340.9231.1
    Tools

    Get Citation

    Copy Citation Text

    Zhaoliang Meng, Zetao Zhang, Yuan Yang, Guofeng Li, Chongbo Tao, Yijiang Niu. Improved XGBoost Stray Current Prediction and Interpretable Model[J]. Laser & Optoelectronics Progress, 2022, 59(12): 1215011

    Download Citation

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

    Category: Machine Vision

    Received: Jul. 16, 2021

    Accepted: Aug. 17, 2021

    Published Online: May. 23, 2022

    The Author Email: Zetao Zhang (zz667892021@163.com)

    DOI:10.3788/LOP202259.1215011

    Topics