Bayesian Neural Ordinary Differential Equations
This work provides a scientific machine learning tool for probabilistic estimation of epistemic uncertainties in Neural ODEs, which is important for researchers and practitioners working with dynamic systems and machine learning models where uncertainty quantification is critical. This is an incremental step in combining existing methods.
This paper explores the integration of Bayesian learning frameworks with Neural Ordinary Differential Equations (NODEs) to quantify uncertainty in model weights. The authors successfully applied MCMC, SGHMC, and SGLD inference methods to NODEs on physical systems and MNIST, achieving 98.5% accuracy on MNIST. They also integrated variational inference with normalizing flows and NODEs, and demonstrated probabilistic identification of model specification in partially-described dynamical systems.
Recently, Neural Ordinary Differential Equations has emerged as a powerful framework for modeling physical simulations without explicitly defining the ODEs governing the system, but instead learning them via machine learning. However, the question: "Can Bayesian learning frameworks be integrated with Neural ODE's to robustly quantify the uncertainty in the weights of a Neural ODE?" remains unanswered. In an effort to address this question, we primarily evaluate the following categories of inference methods: (a) The No-U-Turn MCMC sampler (NUTS), (b) Stochastic Gradient Hamiltonian Monte Carlo (SGHMC) and (c) Stochastic Langevin Gradient Descent (SGLD). We demonstrate the successful integration of Neural ODEs with the above Bayesian inference frameworks on classical physical systems, as well as on standard machine learning datasets like MNIST, using GPU acceleration. On the MNIST dataset, we achieve a posterior sample accuracy of 98.5% on the test ensemble of 10,000 images. Subsequently, for the first time, we demonstrate the successful integration of variational inference with normalizing flows and Neural ODEs, leading to a powerful Bayesian Neural ODE object. Finally, considering a predator-prey model and an epidemiological system, we demonstrate the probabilistic identification of model specification in partially-described dynamical systems using universal ordinary differential equations. Together, this gives a scientific machine learning tool for probabilistic estimation of epistemic uncertainties.