Assoc. Prof. Xuehe Wang

Sun Yat-sen University, China

Biography: Prof. Xuehe Wang obtained her Ph.D. in electrical and electronic engineering from Nanyang Technological University, Singapore in 2016. 

She currently holds the position of Associate Professor at the School of Artificial Intelligence, Sun Yat- sen University, China. Prior to this role, she served as an Assistant Professor at the Infocomm Technology Cluster within Singapore Institute of Technology from 2019 to 2021, and as a postdoctoral research fellow at the Pillar of Engineering Systems and Design, Singapore University of Technology and Design from 2015 to 2019. Her research interests encompass multi-agent systems, federated learning, network economics, and game theory.

Speech Title:  Adaptive federated learning for fast convergence in heterogeneous networks

Abstract:  Federated Learning (FL) has emerged as a prominent distributed machine learning framework that enables geographically discrete clients to train a global model collaboratively while preserving their privacy-sensitive data. However, due to the non-independent-and-identically-distributed (Non-IID) data generated by heterogeneous clients, the performances of the conventional federated optimization schemes such as FedAvg and its variants deteriorate, requiring the design to adaptively adjust specific model parameters to alleviate the negative influence of heterogeneity. In this talk, we will discuss how to design adaptive learning algorithms by leveraging different techniques (such as entropy and norm forms) to alleviate the parameter deviation among heterogeneous clients and achieve fast convergence. The adaptive learning algorithm design requires local information of all clients, including the local parameters and gradients, which is challenging as there is no communication during the local update epochs. To obtain a decentralized adaptive learning rate for each client, we introduce the mean field approach by utilizing mean field terms to estimate the average local parameters and gradients respectively. The simulations show that the proposed framework is superior to the state-of-art FL schemes in both model accuracy and convergent rate on real-world datasets with IID and Non-IID data distribution.

Assoc. Prof. Por Lip Yee

IEEE Senior Member

University of Malaya, Malaysia

Biography: Lip Yee obtained his Ph.D. from the University of Malaya, Malaysia in 2012 under the guidance of Prof. Abdullah bin Gani. Currently, he holds the position of Associate Professor at the Department of System and Computer Technology, Faculty of Computer Science and Information Technology, University of Malaya, Malaysia. He is also a senior member of IEEE. Lip Yee and his team were among the pioneering recipients of grants such as IRPA, E-Science, FRGS, ERGS, PRGS, HIR and IIRG. In 2008, he became the first person to secure two E-Science funds as Principal Investigator (PI). Additionally, he was also the first individual at FCSIT to successfully obtain PRGS and ERGS grants. Apart from collaborating with researchers in Malaysia, Lip Yee has established international collaborations with partners from France UK New Zealand Turkey Thailand China. Furthermore，he has fostered connections with national and international collaborators along with industrial partners in Malaysia and other countries.

Speech Title:  Mitigating Shoulder-Surfing Vulnerabilities in Graphical Authentication

Abstract:This study presents a technical exploration of innovative strategies in graphical authentication aimed at mitigating vulnerabilities associated with shoulder-surfing attacks. In an era where data security is paramount, the development of robust yet user-friendly authentication methods is critical. Traditional alphanumeric passwords are susceptible to brute-force attacks and pose challenges in managing complex combinations. Graphical passwords offer a promising alternative, leveraging human cognitive abilities to recognize and recall images, patterns, and symbols. Our research investigates the potential of graphical passwords, focusing on their resilience against shoulder-surfing threats, wherein malicious actors attempt to capture login credentials by observing legitimate users, posing significant privacy and security risks. We propose techniques to effectively balance security and usability. Furthermore, our empirical analysis demonstrates that our proposed approach not only enhances security but also achieves faster login times compared to conventional methods. This improvement underscores the efficiency of user authentication, making it a practical solution. In summary, this study presents novel advancements in graphical authentication to address shoulder-surfing challenges, reduce cognitive load, and prioritize accessibility and inclusivity in contemporary digital environments.