Build, train, and test a physics informed neural network
Budget: $250 – $750 CAD
I need to model a system based on 6 coupled partial differential equations (PDEs).
I hope to use a physics-informed neural network to model this system. Or you may use physics-constrained machine learning (embedding neural network into the PDE system to approximate some mathematical terms in the PDE system) to model the system.
If possible, please use the "deepXDE" package based on Pytorch backend to build the PINN.
The PINN should be built, trained, and tested. The PINN should reach a satisfactory prediction based on the test set.
I'll give you the training data and more details later on.
I hope to use a physics-informed neural network to model this system. Or you may use physics-constrained machine learning (embedding neural network into the PDE system to approximate some mathematical terms in the PDE system) to model the system.
If possible, please use the "deepXDE" package based on Pytorch backend to build the PINN.
The PINN should be built, trained, and tested. The PINN should reach a satisfactory prediction based on the test set.
I'll give you the training data and more details later on.
Related categories:
Python
Machine Learning (ML)
Neural Networks
Pytorch
Computational Fluid Dynamics