IoT Thrust Seminar | Learning-enabled Networked Systems

In the age of ubiquitous connectivity, networked systems, from data networks to edge/mobile/cloud computing, have reached unprecedented complexity and scale. Traditional methods for network control and resource allocation, often based on static network behavior models, struggle with the dynamic and multifaceted nature of modern networks. Machine learning (ML), with its ability to refine models over time, emerges as a transformative solution, offering potentially vast performance improvements. In this talk, I will introduce my research on developing efficient, scalable, and trustworthy ML algorithms for learning-enabled networked systems, focusing on two critical challenges, scalability and trustworthiness. For scalability, as the size and complexity of networked systems increase, the learning efficiency and computational feasibility substantially decrease. I will present my innovative combinatorial online learning approaches with theoretical guarantees, demonstrating their effectiveness in practical applications to real-world wireless networks and social networks. For trustworthiness, I will delve into the unpredictable behavior of devices in networked systems, showcasing my findings on the vulnerabilities of multi-agent online learning algorithms with concrete examples from recommender systems. Finally, I will discuss the broad potential of ML for redefining network intelligence across various domains, touching upon future challenges and the vision for next-generation networked systems.