Phishing Site Detection Using Logistic Regression and Fine Tuning It Using Various Optimization Parameters

Cybercriminals frequently prefer to deceive people into compromising their own security than using exploits or complex assaults to get access to networks. No matter what technological safeguards are in place, this strategy can be very effective. To conceal their actions and get around conventional security measures, economic resources through phishing attempts might leverage encrypted connections, such as HTTPS. This kind of intrusion has been detected and is being prevented by machine learning technologies. Using two different algorithms (Logistic Regression and Multinomial Naive Bayes) to analyse the URLs and various datasets to compare the results with previous research, we suggested a machine learning-based phishing detection system in this paper. The experimental findings show that the suggested models function superbly and have a high success rate in a social environment.