A Survey on Eliminating Botnet and Intrusion Attacks Through Machine Learning

With the rising linkage and dependence on computer networks, the threat of intrusions and botnets has become a critical concern for organizations and those involved. Malicious actors are in control of unauthorized access or actions that harm the security and integrity of computer systems and insecure devices. The identification of attacks and botnets is of utmost importance in ensuring the security of networks and safeguarding confidential information. In recent years, there has been a growing interest among researchers and organizations in developing unique evaluation methods that can adapt to emerging threats and apply intelligent algorithms such as sparse modelling, transfer learning, distributed computing, or meta-learning capable of detecting, classifying, and predicting previously unknown attack patterns, which can result in machine learning and deep learning models that are more accurate and efficient. This review paper compares the performance of machine learning and deep learning approaches such as Naïve Bayes, Support Vector Machines (SVM), and Logical Regression for intrusion and botnet attacks in various network platforms. This paper provides a comprehensive comparison of machine-learning approaches used in the cyber realm for botnet and intrusion detection and encompasses an analysis of the datasets utilized, the specific cyber issues addressed, and the resultant outcomes. Furthermore, it offers valuable insights into the challenges and advancements within the realm of cybersecurity, thereby facilitating the development of novel approaches and technologies aimed at enhancing digital safety.