The Impact of Duplicate Data on Federated Learning: An Overlooked Issue in Data Quality and Privacy-Preserving Governance
Wei Shao, Associate Research Professor, Shandong Computer Science Center (National Supercomputing Jinan Center), Qilu University of Technology (Shandong Academy of Sciences), China
Abstract:
Federated learning enables collaborative model training across diverse clients without directly sharing raw data,
and has become an important paradigm for privacy-preserving data collaboration in real-world applications. However,
most existing research focuses on model optimization and secure aggregation, while the quality of data before training has
received much less attention.
In real-world scenarios, cross-client data redundancy and overlaps are prevalent due to
cross-platform synchronization or repeated user registration. Such data redundancy is far
from a negligible preprocessing issue and can induce systematic drawbacks to federated
learning pipelines. This talk will first discuss how duplicate data affects federated learning
in terms of training efficiency and model effectiveness, and then further present our
series of privacy-preserving decentralized deduplication solutions for different privacy
and efficiency requirements. These methods enable secure cross-client redundancy
detection while preserving data locality, and provide a practical basis for privacy-aware
data quality governance before training, thereby enhancing the performance of federated
learning.
CV:
Dr. Wei Shao currently works as an associate research professor at Shandong Computer Science Center (National Supercomputer Center in Jinan),
Qilu University of Technology (Shandong Academy of Sciences) in China. She leads and participates in multiple national and provincial research projects
including the National Key R&D Programs, and the National Natural Science Foundation projects. She has published 20+ high-impact papers in top
international journals and conferences, and possesses 10+ national invention patents. Her research focuses on applied cryptography, blockchain security,
privacy-preserving digital identity and trustworthy federated learning. She has proposed multiple innovative schemes including user-controlled anonymous digital identities,
auditable revocable permissioned blockchain mechanisms, and blockchain-based privacy-preserving federated learning frameworks.
Synergies Between Machine Learning and Metaheuristics: New Frontiers in Combinatorial Optimization
Raka Jovanović, Mathematical Institute of the Serbian Academy of Sciences and Arts
Abstract:
Combinatorial optimization problems are central to many scientific and engineering disciplines, yet their NP-hard nature often renders exact
solution methods impractical. Metaheuristics have long provided effective approximate solutions, while recent advances in machine learning
have opened new opportunities to enhance and transform optimization methodologies.
In this talk, I will explore the synergies between these two fields. Machine learning can guide metaheuristic search by learning problem
structures and predicting promising solution regions, while metaheuristics can optimize hyperparameters and neural architectures in machine learning
pipelines. Through case studies from my research — including graph neural networks for combinatorial problems, predictive models for constructive
heuristics, and hybrid approaches in electric vehicle charging, smart agriculture, and logistics — I will illustrate how these synergies can lead
to more efficient and adaptable optimization frameworks. I will also discuss open challenges and future research directions at the intersection
of learning and optimization.
CV:
Dr. Raka Jovanović (born 1978, Belgrade, Serbia) has recently joined the Mathematical Institute of the Serbian Academy of Sciences and Arts as a
Senior Research Associate. He completed his undergraduate, Master's (2007), and Doctoral (2012) studies at the Faculty of Mathematics, University
of Belgrade, under the supervision of Prof. Dr. Milan Tuba. His career includes positions at the Institute of Physics, University of Belgrade (2007-2022)
and as a researcher at Texas A&M University at Qatar (2009-2011). Since 2013, he has been based in Qatar, at the Qatar Environment and Energy Research Institute
(QEERI), part of Hamad Bin Khalifa University in Doha, Qatar where he teaches at the postgraduate level and leads research projects. His primary research directions include
the development of novel metaheuristic methods for hard combinatorial optimization problems, mathematical modeling and optimization of real-world systems in energy and transportation,
and the application of machine learning techniques to complex predictive and decision-making tasks. He has authored over 100 scientific papers published in prestigious
international journals and conference proceedings. He is a recipient of best paper awards, serves as a regular reviewer for top international journals, and maintains
extensive collaborations with institutions in Germany, the United Kingdom, the Netherlands, and others.