Guardians of the Web: Harnessing Machine Learning to Combat Phishing Attacks

Phishing remains one of the most dangerous threats to internet users and organizations today since it utilizes spoofed websites to coax users into revealing their data. This paper focuses on the effectiveness of algorithms in detecting such abusive websites. It goes on to analyze the dataset of phishing and non- phishing URLs providing explanatory attributes such as domain registration date, URL length or the existence of HTTPS. The models studied include Decision Tree, Random Forest, and Support Vector Machines. The results found that the Random Forest algorithm had the best performance of 97% in terms of classification accuracy, and Support Vector Machines performed the best in terms of generalization accuracy with precision and recall values of 0.92 and 0.95, respectively. The study investigates feature selection and determinants of URL structural features which are crucial in determining the efficiency of detection. Also, to enhance model assessment the stratified 10-fold cross-validation technique was performed to reduce bias and variance. These Results show the prospect of One Layer Neural Networks as a tool to improve Phishing Detection Systems and help to provide low-cost and fast solutions for current or future cyberspace struggles. This work aims to increase confidence in online security applications against modern phishing methods.The proposed modifications will help strengthen counter measures against phishing attacks in a shifting technological context while also working towards sustaining the organizations and thus require further inquiry into the facets such as the applicability of sophisticated artificial intelligence techniques the use of useful yet diverse sets of data and the incorporation of explainable intelligent systems