I am a Senior Research Scientist and Early Career Investigator at A*STAR, Singapore. I received my Ph.D. degree from The Hong Kong Polytechnic University in 2017. My research focuses on graph and geometric deep learning, foundation models, and AI for scientific discovery.
My current research develops structure-aware and data-efficient learning methods for foundation models, graph intelligence, and scientific discovery.
My research develops structure-aware, data-efficient, and scientifically grounded learning methods for foundation models, graph intelligence, and AI-enabled scientific discovery.
Long-term research themes that define my broader research agenda.
I study how relational, geometric, and structural priors can be incorporated into modern learning systems. My work includes graph and geometric deep learning, structural attention, sparse computation, and theoretically grounded message passing.
I develop data-efficient and structure-aware approaches for foundation-model adaptation, federated learning, multimodal reasoning, and scalable representation learning.
I apply machine learning and foundation models to scientific problems involving materials, catalysis, biology, environmental systems, weather modelling, and other domains that require structured, multi-fidelity, and physically grounded reasoning.
Current projects that connect methodological research with high-impact applications.
This project develops a next-generation efficient AI framework for accelerating discovery in surface science and catalysis. It brings together representative computational and experimental datasets, multi-fidelity foundation models, sparse learning, and agentic AI to understand realistic surfaces, interfaces, adsorbates, and catalytic environments.
The project aims to move scientific discovery from conventional trial-and-error screening toward an adaptive, autonomous, and self-improving workflow. Key research components include efficient scientific representation learning, multi-fidelity data fusion, hypothesis generation, multi-objective inverse design, uncertainty-aware data acquisition, and experimental validation of promising catalyst candidates.
FLASH develops an AI-enabled decision-support framework for dynamic, zone-based lightning risk management at airports. The project moves beyond uniform, binary lightning warnings by modelling the location, movement, and evolution of lightning systems together with local infrastructure, operational activities, exposure, and human vulnerability.
The framework combines probabilistic lightning nowcasting, spatiotemporal weather modelling, lightning exposure and protection analysis, task-specific hazard assessment, and multimodal risk-data fusion. It will generate continuously updated zone-level risk scores and operational recommendations to support more targeted suspension and resumption decisions while safeguarding airside personnel.
I have had the privilege of working with the following talented researchers and students in my group.
ICASSP 2026 (Area Chair); IJCNN 2025, 2026, and 2027 (Area Chair); EITCE 2025 (Publicity Chair); and ISMIS 2018 (Session Chair).
ICML 2025 and 2026; ICLR 2025 and 2026; NeurIPS 2024 and 2025; AISTATS 2025 and 2026; AAAI 2026; BIBM 2023–2026; WI AI4SG 2021–2025; IJCAI DCM Workshop 2020; and ICTAI 2019.
Associate Editor, Memetic Computing, 2025–Present.
Artificial Intelligence IEEE TPAMI IEEE TKDE IEEE TCYB IEEE TNNLS IEEE TSMC IEEE TFS ACM TKDD IEEE TCBB IEEE TETCI IEEE TNSE IEEE TBD IEEE TCSS Journal of Automation and Intelligence Knowledge and Information Systems Briefings in Bioinformatics Bioinformatics Memetic Computing Neurocomputing Information Sciences
We welcome talented researchers and students interested in advanced machine learning and its scientific applications.
One fully funded postdoctoral position is available in my group for research at the intersection of advanced machine learning and scientific applications.
Applicants with a strong publication record in one or more of these areas are welcome to apply by email with a CV and a brief research statement.
We welcome Ph.D. students and researchers interested in short- or long-term visits supported by national programmes, including the China Scholarship Council, or by research projects from their home institutions or my group.
Prospective visitors may contact me by email with a CV, a brief research statement, and information about the intended funding programme or visit arrangement.