The importance of cybersecurity in contemporary society cannot be inflated,given the substantial impact of networks on various aspects of daily life.Traditional cybersecurity measures,such as anti-virus software and f...The importance of cybersecurity in contemporary society cannot be inflated,given the substantial impact of networks on various aspects of daily life.Traditional cybersecurity measures,such as anti-virus software and firewalls,safeguard networks against potential threats.In network security,using Intrusion Detection Systems(IDSs)is vital for effectively monitoring the various software and hardware components inside a given network.However,they may encounter difficulties when it comes to detecting solitary attacks.Machine Learning(ML)models are implemented in intrusion detection widely because of the high accuracy.The present work aims to assess the performance of machine learning algorithms in the context of intrusion detection,providing valuable insights into their efficacy and potential for enhancing cybersecurity measures.The main objective is to compare the performance of the well-knownML models using the UNSW-NB15 dataset.The performance of the models is discussed in detail with a comparison using evaluation metrics and computational performance.展开更多
文摘The importance of cybersecurity in contemporary society cannot be inflated,given the substantial impact of networks on various aspects of daily life.Traditional cybersecurity measures,such as anti-virus software and firewalls,safeguard networks against potential threats.In network security,using Intrusion Detection Systems(IDSs)is vital for effectively monitoring the various software and hardware components inside a given network.However,they may encounter difficulties when it comes to detecting solitary attacks.Machine Learning(ML)models are implemented in intrusion detection widely because of the high accuracy.The present work aims to assess the performance of machine learning algorithms in the context of intrusion detection,providing valuable insights into their efficacy and potential for enhancing cybersecurity measures.The main objective is to compare the performance of the well-knownML models using the UNSW-NB15 dataset.The performance of the models is discussed in detail with a comparison using evaluation metrics and computational performance.
