Performance Analysis of Machine Learning Algorithms for Password Strength Check

Passwords are a vital component of system security, providing a simple, direct means of protecting a system and representing the identity of an individual for a system. However, the same patterns that people use to create passwords also render them vulnerable to attack. Since the bulk of records has weak passwords or well-known password patterns, password cracking algorithms have been developed to anticipate the password both offline and online. In order To prevent these vulnerabilities from being exploited, it is critical for companies to realize the risks that passwords provide and to create robust policies governing the selection and use of passwords. The proposed model employs multiple Machine Learning (ML) methods, including Decision Tree (DT), Logistic Regression (LR), Random Forest Classifier (RF), and K Nearest Neighbor, pushing users to select a strong password as a protection against these online and offline attacks. The Random Forest Classifier had the best results during the testing of the models over the test set, with an accuracy of 98%, and Logistic Regression achieved the lowest accuracy of 82%.