search

SciML / DataDrivenDiffEq.jl

Data driven modeling and automated discovery of dynamical systems for the SciML Scientific Machine Learning organization
https://docs.sciml.ai/DataDrivenDiffEq/stable/
MIT License
405 stars 57 forks source link

Data driven discovery of PDEs #25

Open finmod opened 4 years ago

finmod commented 4 years ago

https://advances.sciencemag.org/content/advances/suppl/2017/04/24/3.4.e1602614.DC1/1602614_SM.pdf

All the elementary functions being in place, this is a natural extension of DataDrivenDiffEq

AlCap23 commented 4 years ago

Definitely, but as stated in #24, my main focus is getting ISInDy and other methods for ODE and DAE to run.

But once this is done, I can tackle this.

finmod commented 4 years ago

No worry. The sequencing of your workflow may benefit from this approach and steps classification: http://www.dam.brown.edu/people/mmcguirl/DEFD.pdf in developing DataDrivenDiffEq.

The SINDy approach with 4 steps and PDE variant.

The HAVOK approach with 6 steps where Step5 is the Step 4 of SINDy.

finmod commented 4 years ago

@AlCap23 These SINDy and iSINDy are major milestones in learning nonlinear models from datasets. Just a gentle reminder for the extension of the examples to the Lorenz PDE. This would be the cherry on the cake and a fabulous addition to DataDrivenDiffEq.