Related Concept Videos
Propagation of Uncertainty from Systematic Error
Mechanistic Models: Compartment Models in Algorithms for Numerical Problem Solving
In individual population analyses, different algorithms are employed, such as Cauchy's method, which uses a...
Linear Approximation in Time Domain
For a simple pendulum with a mass evenly distributed along its length and the center of mass located at half the pendulum's length,...
Model Approaches for Pharmacokinetic Data: Distributed Parameter Models
The distributed parameter models are specifically designed to account for variations and differences in some drug classes. This model is particularly useful for assessing regional concentrations of anticancer or...
Multi-input and Multi-variable systems
In the absence...
Uncertainty: Overview
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Physics-Informed Neural Networks with Unknown Partial Differential Equations: An Application in Multivariate Time Series.
Variational Bayesian Approximation (VBA): Implementation and Comparison of Different Optimization Algorithms.
Dimensionless Groups by Entropic Similarity: II-Wave Phenomena and Information-Theoretic Flow Regimes.
A Hierarchy of Probability, Fluid and Generalized Densities for the Eulerian Velocivolumetric Description of Fluid Flow, for New Families of Conservation Laws.
Dimensionless Groups by Entropic Similarity: I - Diffusion, Chemical Reaction and Dispersion Processes.
Regularization, Bayesian Inference, and Machine Learning Methods for Inverse Problems.
Topological dependence of viral mutation spread in complex host-interaction networks.
Multifractal signatures of Hamiltonian chaos in Hyperion's rotational dynamics.
Exploring mechanisms for reversal of flow in tunicate hearts.
State estimation in spatiotemporal chaos via low-rank StatFEM.
Universal response functions in driven dissipative tunneling dynamics.
A network-based approach to characterize the dynamics of the coupling field of thermoacoustic oscillators in annular geometry.
Related Experiment Video
Updated: Jun 15, 2025

Development of an Individual-Tree Basal Area Increment Model using a Linear Mixed-Effects Approach
Published on: July 3, 2020
Dynamical system identification, model selection, and model uncertainty quantification by Bayesian inference.
Robert K Niven1, Laurent Cordier2, Ali Mohammad-Djafari3
1School of Engineering and Technology, The University of New South Wales, Canberra, ACT 2600, Australia.
This study introduces a Bayesian framework for identifying dynamical systems from time-series data. This approach offers robust model selection and uncertainty quantification, outperforming traditional sparse regression methods.
Area of Science:
- Dynamical Systems Theory
- Statistical Inference
- Machine Learning
Background:
- Time-series data analysis is crucial for understanding complex systems.
- Traditional methods for dynamical system identification often lack robust uncertainty quantification.
- Sparse regression techniques offer model interpretability but may struggle with complex noise models.
Purpose of the Study:
- To present a Bayesian maximum a posteriori (MAP) framework for dynamical system identification.
- To provide a theoretical justification for regularization terms in system identification.
- To compare Bayesian algorithms with existing sparse regression methods.
Main Methods:
- Developed a Bayesian MAP framework for dynamical system identification.
- Equated the framework to generalized Tikhonov regularization.
- Employed joint MAP and variational Bayesian approximation algorithms.
- Compared performance against LASSO, ridge regression, and SINDy algorithms.
Main Results:
- The Bayesian framework provides a rational basis for residual and regularization terms.
- Bayesian inference allows for model ranking, uncertainty quantification, and hyperparameter estimation.
- The posterior Gaussian norm serves as a robust metric for quantitative model selection.
- Bayesian methods demonstrated superior performance in identifying dynamical systems with various noise types.
Conclusions:
- The proposed Bayesian MAP framework offers a principled approach to dynamical system identification.
- It provides enhanced capabilities for model selection and uncertainty quantification compared to existing methods.
- The framework is particularly effective for systems with Gaussian or Laplace noise.

