基于TF-IDF加权朴素贝叶斯算法的ATP车载设备测试案例分类研究  

Classification of ATP on-board equipment test cases based on TF-IDF weighted Naive Bayesian algorithm

作　　者：王心仪 程剑锋 刘育君 WANG Xinyi;CHENG Jianfeng;LIU Yujun(Postgraduate Department,China Academy of Railway Sciences,Beijing 100081,China;Signal&Communication Research Institute,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China)

机构地区：[1]中国铁道科学研究院研究生部,北京100081 [2]中国铁道科学研究院集团有限公司通信信号研究所,北京100081

出　　处：《铁路计算机应用》2022年第12期8-12,共5页Railway Computer Application

基　　金：中国铁道科学研究院集团有限公司科研项目(2021YJ085);北京华铁信息技术有限公司科研项目(2019HT22)。

摘　　要：针对列车超速防护(ATP,Automatic Train Protection)系统车载设备测试案例分类存在的工作量大、效率低且准确性不高等问题,提出了将词频—逆文档频率(TF-IDF,Term Frequency-Inverse Document Frequency)与朴素贝叶斯算法相结合,应用于测试案例分类的方案。利用TF-IDF算法筛选特征词及权重,对朴素贝叶斯算法进行加权处理,并基于实验室现有ATP车载设备的测试案例进行验证。实验结果表明,文章的特征词提取及测试案例分类方法具有较高的准确性。Aiming at the problems of heavy workload, low efficiency and low accuracy in the classification of test cases of on-board equipment of ATP(Automatic Train Protection) system, this paper proposed a scheme that combined TF-IDF(Term Frequency Inverse Document Frequency) with Naive Bayesian algorithm to classify test cases. The paper used TF-IDF algorithm to filter feature words and weights, and weighted Naive Bayesian algorithm, which was verified based on the test cases of existing ATP on-board equipment in the laboratory. The experiment results show that the method of feature word extraction and test case classification has high accuracy.

关 键 词：列车超速防护(ATP) 测试案例 TF-IDF 朴素贝叶斯 案例分类 

分 类 号：U284.48[交通运输工程—交通信息工程及控制] TP39[交通运输工程—道路与铁道工程]