Mandujano Reyes, J. F., T. F. Ma, I. P. McGahan, D. J. Storm, D. P. Walsh and J. Zhu. 2023. Spatio-temporal ecological models via physics-informed neural net-
works for studying chronic wasting disease. Spatial Statistics 62:100850. https://doi.org/10.1016/j.spasta.2024.100850
Abstract
To mitigate the negative effects of emerging wildlife diseases in biodiver-
sity and public health it is critical to accurately forecast pathogen dissemi-
nation while incorporating relevant spatio-temporal covariates. Forecasting
spatio-temporal processes can often be improved by incorporating scientific
knowledge about the dynamics of the process using physical models. Eco-
logical diffusion equations are often used to model epidemiological processes
of wildlife diseases where environmental factors play a role in disease spread.
Physics-informed neural networks (PINN) are deep learning algorithms that
constrain neural network predictions based on physical laws and therefore
are powerful forecasting models useful even in cases of limited and imperfect
training data. In this paper, we develop a novel ecological modeling tool
using PINNs, which fits a feedforward neural network and simultaneously
performs parameter identification in a partial differential equation (PDE)
with varying coefficients. We demonstrate the applicability of our model by
comparing it with the commonly used Bayesian stochastic partial differential
equation method and traditional machine learning approaches, showing that
our proposed model exhibits superior prediction and forecasting performance when modeling chronic wasting disease in deer in Wisconsin. Furthermore,
our model provides the opportunity to obtain scientific insights about spatio-
temporal covariates affecting spread and growth of diseases. This work con-
tributes to future machine learning and statistical developments with the
objective of studying spatio-temporal processes enhanced by prior physical
knowledge.
sity and public health it is critical to accurately forecast pathogen dissemi-
nation while incorporating relevant spatio-temporal covariates. Forecasting
spatio-temporal processes can often be improved by incorporating scientific
knowledge about the dynamics of the process using physical models. Eco-
logical diffusion equations are often used to model epidemiological processes
of wildlife diseases where environmental factors play a role in disease spread.
Physics-informed neural networks (PINN) are deep learning algorithms that
constrain neural network predictions based on physical laws and therefore
are powerful forecasting models useful even in cases of limited and imperfect
training data. In this paper, we develop a novel ecological modeling tool
using PINNs, which fits a feedforward neural network and simultaneously
performs parameter identification in a partial differential equation (PDE)
with varying coefficients. We demonstrate the applicability of our model by
comparing it with the commonly used Bayesian stochastic partial differential
equation method and traditional machine learning approaches, showing that
our proposed model exhibits superior prediction and forecasting performance when modeling chronic wasting disease in deer in Wisconsin. Furthermore,
our model provides the opportunity to obtain scientific insights about spatio-
temporal covariates affecting spread and growth of diseases. This work con-
tributes to future machine learning and statistical developments with the
objective of studying spatio-temporal processes enhanced by prior physical
knowledge.