S M Asiful Huda

I am a Marie Skłodowska-Curie doctoral candidate at the University of Naples Federico II (UNINA), Italy, in the MSCA Doctoral Network TUAI (Towards an Understanding of Artificial Intelligence). My doctoral project, DC9: Decentralised Graph Neural Networks: Adaptation, Training and Interpretability in Federated Environments, is supervised by Prof. Francesco Piccialli (UNINA), with co-supervisors Prof. David Camacho (UPM), Prof. Shen Yin (NTNU) and Prof. Dariusz Mrozek (SUT).

My long-term research goal is to advance trustworthy AI systems that can both learn and unlearn efficiently while remaining robust and reliable. As AI models increasingly rely on large-scale data, they must also be able to forget specific information when necessary, whether due to privacy regulations (e.g., GDPR), harmful or copyrighted content, or poisoned and outdated data. These challenges become particularly complex in decentralised and relational settings, where data is distributed across institutions (e.g., hospitals and energy providers) and interconnected as graphs. I am interested in enabling collaborative learning and unlearning while ensuring transparency, accountability, and reliability.

My core research areas include:

• Federated and Decentralised Learning — collaborative training across heterogeneous clients, aggregation strategies, and learning under non-IID data
• Graph Neural Networks — federated GNNs, learning over distributed graph-structured data, and interpretability of graph models
• Machine Unlearning — efficient and verifiable removal of data from trained models, especially in federated and graph settings
• Trustworthy and Privacy-Preserving AI — explainability, robustness, and privacy guarantees for decentralised systems
• Applications in healthcare, smart grids, communication networks, and edge computing

Before enrolling into PhD, I was a Research Associate at the Gwangju Institute of Science and Technology (GIST), South Korea, working on machine learning for wireless power transfer and electricity price forecasting, and a Researcher at HT Core Co., Ltd., developing an XR-based training platform for clinical practice. I received my M.E. in Computer Engineering from Chosun University, South Korea, in 2023, where I worked on reinforcement learning-based computation offloading for UAV-aided edge computing under the supervision of Prof. Sangman Moh. I received my B.Sc. in Computer Science from East West University, Bangladesh.