MACHINE LEARNING-POWERED CYBERSECURITY: DETECTING PHISHING AND RANSOMWARE IN FINANCIAL INSTITUTIONS

Phishing and ransomware attacks are significant cyber threats that target financial institutions, aiming to deceive users and exploit vulnerabilities for malicious gains. Phishing attacks often involve fraudulent emails or websites that trick users into revealing sensitive information, while ransomware encrypts a victim's data and demands payment for its release. Both types of attacks pose severe risks to financial institutions, potentially leading to data breaches, financial losses, and reputational damage. To combat these threats, this paper proposed an advanced detection system using machine learning techniques. The proposed system focused on feature engineering and training a Random Forest classifier to detect phishing and ransomware attacks based on key attributes like URL structure and file characteristics. For phishing detection, features such as URL length, subdomains, and the presence of secure HTTPS protocols were extracted, while for ransomware detection, file name length, the presence of executable extensions, and suspicious keywords were analyzed. The results of the proposed system showed a significant improvement over existing systems, achieving an accuracy of 99.61%.