基于S-XGBoost融合模型的中国进境快件风险水平评价研究
Risk Level of China's Inbound Express Delivery Based on S-XGBoost Fusion Model Evaluation Research
刘昌伟 1谢晶 2邹廉青 3梅云鹏 4Kashif Abbass5
作者信息
- 1. 大连海关所属大连港湾海关办公室
- 2. 上海海关学院海关与公共管理学院
- 3. 大连海关风险防控分局风险分析三科
- 4. 大连海关所属金普海关查检一科
- 5. School of Economics and Management Nanjing University of Science and Technology
- 折叠
摘要
党的二十大报告指出要增强维护国家安全能力,推进国家安全体系和能力现代化,坚决维护国家安全和社会稳定.随着互联网产业的发展壮大,以进境快件为代表的跨境个人消费进入发展的快车道,随之而来的安全准入风险也越来越高.为深入推进智慧海关建设,进一步提升进境快件风险防控能力,文章分析了现行监管模式下进境快件存在的问题,运用XGBoost算法和SMOTE技术构建了S-XGBoost进境快件风险评估模型,并对历史通关数据进行了仿真布控和评估.研究结果表明本文提出的S-XGBoost算法模型具有较好的预测效果,风险预测准确率达到了 99.06%,能够为海关工作人员提升进境快件布控有效率提出切实参考建议.
Abstract
The report of the 20th National Congress of the Communist Party of China pointed out the need to enhance the ability to maintain national security,promote the modernization of the national security system and capabilities,and resolutely maintain national security and social stability.With the development and growth of the internet industry,cross-border personal consumption represented by in-bound express has entered the fast lane of development,and the accompanying security risks at the bor-der are also increasing.In order to further promote the construction of smart customs and enhance the risk prevention and control capabilities of inbound express shipments,the paper analyzes the problems existing in the current regulatory model of inbound express shipments,constructs an S-XGBoost inbound express risk assessment model using XGBoost algorithm and SMOTE technology,and conducts simulation control and evaluation of historical customs clearance data.The research results indicate that the S-XGBoost algorithm model proposed in this paper has good predictive performance,with a risk pre-diction accuracy of 99.06%.It can provide practical reference suggestions for customs staff to improve the efficiency of inbound express delivery control.
关键词
数据挖掘/中国进境快件/S-XGBoost/风险水平Key words
Data Mining/China's Inbound Express/S-XGBoost/Risk Level引用本文复制引用
基金项目
上海海关学院研究生科研创新项目(2316047A2022)
出版年
2024
NETL
NSTL
万方数据