by Da Long, Zheng Wang, Aditi Krishnapriyan, Robert Kirby, Shandian Zhe, Michael Mahoney
Physical modeling is critical for many modern science and engineering applications. From a data science or machine learning perspective, where more domain-agnostic, data-driven models are pervasive, physical knowledge — often expressed as differential equations — is valuable in that it is complementary to data, and it can potentially help overcome issues such as data sparsity, noise, and inaccuracy. In this work, we propose a simple, yet powerful and general framework — AutoIP, for Automatically Incorporating Physics — that can integrate all kinds of differential equations into Gaussian Processes (GPs) to enhance prediction accuracy and uncertainty quantification. These equations can be linear or nonlinear, spatial, temporal, or spatio-temporal, complete or incomplete with unknown source terms, and so on. Based on kernel differentiation, we construct a GP prior to sample the values of the target function, equation related derivatives, and latent source functions, which are all jointly from a multivariate Gaussian distribution. The sampled values are fed to two likelihoods: one to fit the observations, and the other to conform to the equation. We use the whitening method to evade the strong dependency between the sample.
We implemented our model by Jax and pytorch. For Jax version, Jax and Optax --- A Jax optimizaiton library are required. For pytorch version, pytorch is required.
- No damping:
./Jax_version/no_damping.py or ./pytorch_version/no_damping.py - With latent source + no damping:
./pytorch_version/latent_no_damping.py - Damping:
./Jax_version/damping.py or ./pytorch_version/damping.py - With latent source + damping:
./pytorch_version/latent_damping.py
To run noisy pendulum, add "_noise" to the end. For example, to run noisy damping pendulum: use ./Jax_version/damping_noise.py or ./pytorch_version/damping_noise.py.
- Full equation: ./Jax_version/allen.py or ./pytorch_version/allen.py
- With latent source: ./pytorch_version/latent_allen.py
IFC is released under the MIT License, please refer the LICENSE for details
Feel free to contact us via dl932@cs.utah.edu or u1368737@utah.edu
Please cite our paper if it is helpful to you
@inproceedings{long2022autoip,
title={AutoIP: A United Framework to Integrate Physics into Gaussian Processes},
author={Long, Da and Wang, Zheng and Krishnapriyan, Aditi and Kirby, Robert and Zhe, Shandian and Mahoney, Michael},
booktitle={International Conference on Machine Learning},
pages={14210--14222},
year={2022},
organization={PMLR}
}