Security and Machine Learning
Research Problem: In the rapidly evolving digital landscape, cybersecurity threats continue to grow in both complexity and frequency. It poses significant challenges to the protection of sensitive data and critical infrastructures. Despite advancements in security protocols, traditional methods often struggle to keep pace with the adaptive nature of cyber-attacks, such as malware, phishing, and distributed denial-of-service (DDoS) attacks. Machine learning (ML) has emerged as a promising solution due to its ability to identify patterns, predict threats, and automate response mechanisms. However, there are significant research gaps in the effective integration of ML techniques into security frameworks. These challenges include managing the imbalance in labeled data for training, reducing false positives, ensuring model robustness against adversarial attacks, and maintaining the interpretability and transparency of ML-based systems. The problem, therefore, is how to design, develop, and implement machine learning algorithms that enhance security frameworks by providing accurate, real-time detection and response to cyber threats while addressing the issues of data imbalance, adversarial resilience, and explainability.
Security and Blockchain
Research problem: In centralized systems, privacy of user identity information relies on centralized entity, such as system administrator. Since centralized entity is responsible to distribute keys to all the users in the system, he can read and modify the identity information. Compromise of such centralized entity results into the entire system vulnerability. Therefore, centralized systems often failed to provide the data privacy and integrity. In addition, centralized systems are also vulnerable to single-point-of-failure. Blockchain based solution is a possible alternate to provide integrity and protect system from single-point-of-failure. But, the ledger of the blockchain, contains all the transactions information, is distributed and available to each member of the blockchain network. Hence, blockchain-based identity management for user authentication, while preserving privacy, is a challenging problem.