Automatic Extraction of Ordinary Differential Equations from Data: Sparse Regression Tools for System Identification
Published in Durham University, 2023
This thesis develops three strategies for automatically identifying ordinary differential equation models from data. The methods combine signal filtering, sparse regression, bootstrap sampling, Bayesian inference, and unsupervised learning to recover nonlinear dynamics reliably. Each is benchmarked against state-of-the-art system identification algorithms, with applications including weather forecasting, chemical reaction and electrical circuit modelling, and predator-prey dynamics.
Recommended citation: Egan, K. (2023). "Automatic Extraction of Ordinary Differential Equations from Data: Sparse Regression Tools for System Identification." PhD thesis, Durham University.
Download Paper