Detecting Brute Force Attacks Using Machine Learning

Abstract

The importance of identifying network traffic abnormalities in cybersecurity cannot be emphasized enough, particularly considering the growing frequency and complexity of computer network assaults. With the emergence of new Internet-related technology, there is a corresponding increase in complex assaults. A significant difficulty is dictionary-based brute-force assaults (BFA), which need efficient real-time detection and mitigation techniques. This study explores the detection of SSH and FTP brute-force attacks via the use of the primary objective of our study is to use the machine learning methodology for the identification and detection of SSH and FTP brute-force assaults. Employing many classifiers enables a pretty comprehensive examination of the effectiveness of machine learners in spotting brute force attacks on SSH and FTP. Brute-force attacks are a widely used and perilous technique for acquiring usernames and passwords. Utilizing ethical hacking is a commendable method to assess the impact of a brute-force attack. This article explores several defense tactics and methodologies for using brute-force attacks. The advantages and disadvantages of many defense techniques are listed, along with details on the kind of attack that is most straightforward to recognize. we made use of machine learning (ML) classifiers: Naive Bayes (NB), decision Tree (DT), random forest (RF) Logistic Regression (LG), Quadratic Discriminant Analysis (QDA), Stochastic Gradient Descent (SGD), Linear Discriminant Analysis (LDA), Multi-Layer Perceptron (MLP), we determined that the Random Forest (RF) algorithm achieved the highest level with an accuracy of 99.9%.