mit-mseas/neuralClosureModels

Code for the framework, neural closure models.

25
/ 100
Experimental

This project helps researchers and scientists improve predictions for complex dynamical systems by learning how to account for neglected details in simplified models. It takes data from high-fidelity simulations to produce 'neural closure models' that enhance the accuracy of predictions from faster, lower-fidelity models. This is ideal for scientists, engineers, and researchers working with physical, biological, or environmental simulations where computational cost is a barrier to using full-complexity models.

No commits in the last 6 months.

Use this if you need to make accurate long-term predictions from simplified or coarse models of complex systems, and you have access to data from high-fidelity simulations.

Not ideal if you do not work with dynamical systems, do not have high-fidelity simulation data, or are not concerned with the computational efficiency of your models.

dynamical-systems scientific-modeling computational-physics biogeochemical-modeling predictive-modeling
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 4 / 25
Maturity 8 / 25
Community 13 / 25

How are scores calculated?

Stars

8

Forks

2

Language

Jupyter Notebook

License

Last pushed

May 29, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/mit-mseas/neuralClosureModels"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.