Prof. Zhang Shijun, Assistant Professor of the Department of Applied Mathematics at The Hong Kong Polytechnic University (PolyU), has been honoured with the prestigious 2026 Frontiers of Science Award presented by the International Congress of Basic Science (ICBS) in recognition of his outstanding research, which demonstrates profound impact and forward-looking vision.
Prof. Zhang was recognised in the mathematics (numerical analysis) category for his paper, "Deep network approximation for smooth functions," published in the SIAM Journal on Mathematical Analysis (2021). He shares this distinction with co-authors from Duke University, the National University of Singapore, and the University of Maryland.
Prof. Zhang’s research focuses on advancing the theoretical foundations and practical capabilities of deep neural networks. His work examines how neural networks approximate complex functions, aiming to minimise critical approximation, generalisation, and optimisation errors, thereby bridging the gap between theory and application.
Beyond these core principles, his research spans neural network optimisation, activation function properties, and the expressive power of modern architectures such as Transformers and Graph Neural Networks. He also explores cutting-edge techniques, including transfer learning, neural architecture search, and model compression, to drive next-generation innovations in artificial intelligence.
The Frontiers of Science Award recognises exceptional, original research published over the past decade across Mathematics, Physics, and Information Science & Engineering. To receive this honour, a scientific achievement must be of the highest scholarly value and have made a major global impact in its field.
