摘要
文章基于用户的移动通信数据进行诈骗用户特征识别,采用XGBoost、LightGBM等有监督的机器学习算法及多模型融合技术实现疑似电信诈骗行为发现和欺诈用户清单输出,将欺诈用户清单反馈给运营商,由运营商做出预警处理,及时提醒用户,防止电信诈骗事件发生。该防欺诈模型在验证集上进行了验证,提高了防欺诈模型的准确率。
This paper identifies the characteristics of fraudulent users based on mobile communications data,and uses supervising machine learning algorithms such as XGBoost and LightGBM and multi-model fu⁃sion technology to detect suspected telecom frauds.The model can output the list of fraudulent users finally and feed the list of fraudulent users back to the operators.The operators will make early warning treatment to remind their users in time to prevent telecom frauds.In this paper,the anti-fraud model is verified on the verification sets,which improves the accuracy of the anti-fraud model.
作者
毕佳佳
李京文
BI Jia-jia;LI Jing-wen
出处
《安徽职业技术学院学报》
2021年第4期11-16,共6页
Journal of Anhui Vocational & Technical College
基金
2020年安徽省自然科学项目“大数据技术在大气污染分析及防治中应用”(KJ2020A1040)。
关键词
防诈骗
机器学习
多模型融合
anti-fraud
machine learning
multi-model fusion