The invention discloses a traffic abnormal data anomaly detection method based on an improved robust random forest. The method comprises the following steps: S1, data input; s2, building a model; s3, data processing; s4, abnormal score calculation; s5, determining a threshold value; s6, performing anomaly detection; and S7, application. S8, evaluating and optimizing the model; and S9, optimizing the overall scheme. According to the method, the weight is adaptively allocated to each tree, and the performance of the trees is fully considered, so that the accuracy and robustness of the whole model are improved. By introducing a robustness enhancement mechanism for abnormal values and a self-adaptive parameter optimization strategy, the algorithm can better adapt to different traffic scenes and data distribution, and the generalization performance of the algorithm is improved.
本发明公开了基于改进强健随机森林的交通异常数据异常检测方法,包括如下步骤:S1、数据输入;S2、模型搭建;S3、数据处理;S4、异常分数计算;S5、阈值确定;S6、异常检测;S7、应用;S8、模型评估和优化;S9、优化总体方案。本发明通过自适应地为每颗树分配权重,充分考虑树的性能,从而提高整体模型的准确性和鲁棒性。通过引入对异常值的鲁棒性增强机制和自适应的参数优化策略,算法能够更好地适应不同的交通场景和数据分布,提高了算法的泛化性能。
Traffic abnormal data anomaly detection method based on improved robust random forest
基于改进强健随机森林的交通异常数据异常检测方法
2024-04-12
Patent
Electronic Resource
Chinese
Adaptive traffic event detection method based on improved random forest
European Patent Office | 2021
|European Patent Office | 2020
|European Patent Office | 2023
|