一種基于優(yōu)化機器學習模型的網(wǎng)絡入侵檢測分類方法
打開文本圖片集
doi：10.3969/J.ISSN.1672-7274.2025.06.007
中圖分類號：TP181；TP309 文獻標志碼：A 文章編碼：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 特征工程設計
在互聯(lián)網(wǎng)快速發(fā)展的同時，網(wǎng)絡安全問題也日益嚴重，網(wǎng)絡入侵行為頻發(fā)，給個人、企業(yè)和國家?guī)砹司薮蟮膿p失和風險。（剩余3972字）