ENHANCING CYBERSECURITY IN CLOUD COMPUTING: A NOVEL APPROACH USING BLOCKCHAIN AND MACHINE LEARNING

With the fast growth of cloud computing, there are now new and changing cybersecurity risks that require smart and easily monitored defenses. In such settings, federated learning and privacy-preserving machine learning have been successful because it is hard to bring all the data together. In this study, a new approach is introduced that combines machine learning for threat detection and blockchain-style audit logging to improve security and responsibility in cloud environments. Using two real-world datasets, we trained a Random Forest classifier that was able to correctly detect every malicious domain by analyzing antivirus reports and reputation scores. In addition, a model for attributing threats to groups of threat actors achieved only moderate results, showing how difficult it is to do attribution with structured data. We designed a blockchain simulation that adds cryptographic hashes to each machine learning prediction which allows for verifiable review. Evaluating worldwide threat data together demonstrated that DDoS, ransomware and zero-day vulnerabilities are increasing, making it even more important to have predictive and transparent cloud defense. This new method combines ML and blockchain to ensure both accurate detection and compliance which is essential for modern cloud-based environments. Future improvements will involve running on live blockchain networks, linking with federated learning methods for privacy-focused edge-cloud security and adjusting for real-time analysis of cloud traffic..