Staff profile

Weizhen Li is a PhD student at the Department of Engineering, supported by Durham Engineering Doctoral Studentship. He received MSc in Statistics with Data Science from the University of Edinburgh in 2019 and MSc in Geoinformation Technology and Cartography from the University of Glasgow in 2020. Before this, he got the BEc in Economic Statistics at Hainan University, China. He is supervised by Dr Rui Carvalho and Professor Hongjian Sun. His research interests include computational mathematics, signal processing, regression analysis and machine learning. 

Research Project

My research project is about identifying governing equations from dynamic systems via statistical methods. Since the rapid development of computers, we have the sufficient ability to derive the governing equations hidden in the phenomenon. The identification can be achieved by analysing the collected data via data-driven methods. The current data-driven algorithms still need prior knowledge to set the appropriate parameters. Under the proper setting, these algorithms can identify governing equations close to the truth; otherwise, they make lots of bias. Moreover, some deep learning methods can train the tuning parameters automatically, but they are computationally expensive, always taking a long time in training parameters. We need a data-driven and computationally cheap method that can be applied to various engineering problems without expert knowledge. 

Journal Article

• Egan, K., Li, W., & Carvalho, R. (2024). Automatically discovering ordinary differential equations from data with sparse regression. Communications Physics, 7(1), Article 20. https://doi.org/10.1038/s42005-023-01516-2

Report

• Egan, K., Li, W., & Carvalho, R. (2023). Automatically identifying ordinary differential equations from data. Durham University