一种基于优化机器学习模型的网络入侵检测分类方法

打开文本图片集

doi：10.3969/J.ISSN.1672-7274.2025.06.007

中图分类号：TP181；TP309 文献标志码：A 文章编码：1672-7274（2025）06-0019-03

A Network Intrusion Detection Classification Method Based on Optimized Machine Learning Models

YUZhenghong，HE Zhaoyin²，LIU Shuaifu³ （1. Inspur Electronic Information Industry Co.， Ltd.， Ji'nan 2501oo， China; 2. Inspur Group Co.， Ltd.，Ji'nan 25010o，China;3.Ji'nan InspurData Technology Co.，Ltd.，Ji'nan250100，China）

Abstract： With the rapid development of the Internet， network security problems are becoming increasingly serious，and network intrusionsocur frequently，posing a huge threat to individuals，enterprises andcountries. Therefore，researching and developing efficientandaccuratemethods fornetwork intrusiondetection isof great significance.Therefore，a network intrusion detection clasification method based on optimized machine learning models is proposed.This method extracts key features from network trafcdata through featureengineering，performs data preprocessing to ensure data quality， comprehensively evaluates model performance using validation datasets， and continuously improves model stability and accuracy through adjusting features，optimizing parameters，and ensemble learning methods，achieving automated detection and real-time monitoring.This method can efectively identify network intrusion behaviors，improve network security protection capabilities，and provide strong support for building a more secure and reliable network environment.

Keywords： network intrusion detection; machine learning; feature engineering

1 特征工程设计

在互联网快速发展的同时，网络安全问题也日益严重，网络入侵行为频发，给个人、企业和国家带来了巨大的损失和风险。（剩余3972字）