Intrusion detection system based on machine learning
Cybersecurity has become a globally significant issue.Intrusion Detection Systems(IDS)play a crucial role in safeguarding interconnected networks by detecting malicious actors and activities.However,as the volume of data grows,dimen-sionality reduction becomes increasingly challenging during ML model training.Addressing this,the article introduces a novel ML-based network intrusion detection model.This model utilizes Random Oversampling(RO)to tackle data imbalance,incorpo-rates feature embedding based on clustering results,and employs Principal Component Analysis(PCA)for dimensionality reduc-tion,tailored for large and imbalanced datasets.The model performs remarkably well on the UNSW-NB15 dataset,with RF and ET models achieving accuracies of 99.53%and 99.14%respectively,surpassing existing technologies.It can accurately monitor and identify network traffic intrusions,ensuring network security.
万方数据
