Tara Salman

Research Interest 

My current research lies in the intersection of robust, secure, and distributed artificial intelligence, Blockchains (or scalable distributed systems), and security applications. Specifically, the research combines blockchains and AI to build scalable cybersecurity solutions. 

My current research focuses on distributing and securing the learning processing where the data is globally distributed. Specific applications include building robust and explainable federated learning under adverserial setting, attack landscape of new large language models, efficient and secure heirarchal federated learning, efficient blockchain-based federated learning and blockchain-based reinforcement learning, and their novel cybersecurity applications. Aside from that, I work on machine learning/deep learning applications to cybersecurity especially in the IoT domains, blockchains' performance under adversarial settings, and blockchains in the quantum era. 

My Ph.D. research was on extending blockchains from simple storage systems to knowledge systems. The stored blockchain data includes individual decisions, opinions, or responses, and a global decision is to be made within the blockchain process. This research has been applied to intrusion detection, malware detection, and fintech decision-making applications. However, the work so far has been on combining AI predictions within the blockchain process, assuming that AI models are already deploying. 

The full list of publications can be found here and through my Google scholar.