Scholarlink

Train delays have a big impact on customer happiness and railway management. Predicting train delays accurately may enhance both passenger satisfaction and operational effectiveness. Although many other algorithms have been used to solve this issue, Support Vector Machines (SVM) in conjunction with strong preprocessing methods may provide better results. This study investigates how several preprocessing methods, namely the Modified Z-score method (MZ-Score), Interquartile Range (IQR) method, and Z-score method, affect the effectiveness of SVM in predicting train delays. This study aims to improve train delay prediction models’ recall, accuracy, and precision by using powerful preprocessing methods in conjunction with SVM. Testing the suggested system using the Train_delay_Prediction.csv dataset shows that combining SVM with the MZ-score technique yields better results in terms of accuracy, precision, and recall.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Efficient Train Traffic Prediction Using Support Vector Machines with Advanced Preprocessing Techniques

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • C. Radhika,
  • D. Kerana Hanirex

摘要

Train delays have a big impact on customer happiness and railway management. Predicting train delays accurately may enhance both passenger satisfaction and operational effectiveness. Although many other algorithms have been used to solve this issue, Support Vector Machines (SVM) in conjunction with strong preprocessing methods may provide better results. This study investigates how several preprocessing methods, namely the Modified Z-score method (MZ-Score), Interquartile Range (IQR) method, and Z-score method, affect the effectiveness of SVM in predicting train delays. This study aims to improve train delay prediction models’ recall, accuracy, and precision by using powerful preprocessing methods in conjunction with SVM. Testing the suggested system using the Train_delay_Prediction.csv dataset shows that combining SVM with the MZ-score technique yields better results in terms of accuracy, precision, and recall.