Scholarlink

This paper focuses on detecting network attacks using anomaly-based identification systems. The research introduces robust machine learning techniques to identify network intrusion detection including its type using the dataset, namely, UNSW-NB15. It covers nine distinct attacks with the help of 49 features. The decision tree classifier has given the best result and noted accuracy is 99.05% among other techniques, namely, random forest, Adaboost, XGboost, KNN, and SVM. The study emphasized the relevance and strength of all features in network attack detection, obviating the need for feature selection.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning-Based Network Attacks Detection

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Muthukumarapandian Chandrasekaran,
  • Chetana Tailor

摘要

This paper focuses on detecting network attacks using anomaly-based identification systems. The research introduces robust machine learning techniques to identify network intrusion detection including its type using the dataset, namely, UNSW-NB15. It covers nine distinct attacks with the help of 49 features. The decision tree classifier has given the best result and noted accuracy is 99.05% among other techniques, namely, random forest, Adaboost, XGboost, KNN, and SVM. The study emphasized the relevance and strength of all features in network attack detection, obviating the need for feature selection.