The invention discloses a traffic violation model training process anomaly detection method, and particularly relates to the field of traffic violation detection, and the method specifically comprises the steps: S1, preparing a data set for training a traffic violation model, and carrying out the preprocessing and cleaning of data; s2, extracting related features from the original data according to the characteristics of the traffic violation prediction task; s3, training a traffic violation prediction model by using the extracted features and the marked data; s4, in the model training process, detecting whether an abnormal sample or a model error exists or not by using an anomaly detection algorithm; s5, if abnormal samples or model errors are detected, corresponding measures need to be taken for processing; s6, after training is completed, the test data set is used for evaluating the model, and whether the prediction performance of the model meets the requirement or not is checked; and S7, deploying the trained model to practical application. The method improves the model performance, avoids overfitting, optimizes the model deployment, reduces the misjudgment rate, and is suitable for traffic violation detection.

    本发明公开了一种交通违法模型训练过程异常检测方法,具体涉及交通违法检测领域,该方法具体步骤为:S1:准备用于训练交通违法模型的数据集,并对数据进行预处理和清洗;S2:根据交通违法预测任务的特点,从原始数据中提取相关特征;S3:使用提取的特征和标注好的数据,训练交通违法预测模型;S4:在模型训练过程中,使用异常检测算法来检测是否存在异常样本或模型错误;S5:如果检测到异常样本或模型错误,需要采取相应的措施来处理;S6:在训练完成后,使用测试数据集对模型进行评估,检查模型的预测性能是否满足要求;S7:将训练好的模型部署到实际应用中。本发明提高模型性能、避免过拟合、优化模型部署、降低误判率,适用于交通违法检测。


    Access

    Download


    Export, share and cite



    Title :

    Traffic violation model training process anomaly detection method


    Additional title:

    一种交通违法模型训练过程异常检测方法


    Contributors:
    SUN XIAOLE (author) / XIA BOCHENG (author) / WANG YINXIA (author)

    Publication date :

    2024-03-05


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS



    TRAFFIC VIOLATION DETECTION DEVICE, SYSTEM, TRAFFIC VIOLATION DETECTION METHOD, AND PROGRAM

    HARADA NAOYUKI / UMETANI HIDEO / MATSUI SHUHEI et al. | European Patent Office | 2018

    Free access

    Traffic violation detection

    BULAN ORHAN / BURRY AARON M / LOCE ROBERT P | European Patent Office | 2017

    Free access

    Traffic anomaly detection method and device and model training method and device

    YE JIN / TAN XIAO / SUN HAO | European Patent Office | 2021

    Free access

    Vehicle model filtering-based traffic violation detection method

    CAO QUAN / WANG YU | European Patent Office | 2015

    Free access

    Intelligent Traffic Violation Detection

    Ravish, Roopa / Rangaswamy, Shanta / Char, Kausthub | IEEE | 2021