SSH ATTACKS DETECTION USING MACHINE LEARNING: COMPARATIVE ANALYSIS OF DIFFERENT ML MODELS

Abstract

Given the ubiquity of SSH as one of most famous communication protocol, SSH attacks can be detrimental. Once, the attacker is able to guess the credentials using brute force attack, they can compromise sensitive information, bring down a business, and whatnot; the realm of possibilities is unimaginable. Therefore, there must be some way of detecting SSH attacks so that the administration might prevent these attacks before the attacker can find the credentials. Hence, this project aims to make a comprehensive study of some state of the art machine learning algorithms used for SSH detection and, in turn, compare the results of these algorithms.