Simulation By Data ONLY: Fourier Neural Operator (FNO)

Please see our courses in the following link:
www.courses.machinedecision.com/
Please visit our website in the following link:
www.machinedecision.com/
Physics Informed Neural Networks (PINNs)
Inverse Physics Informed Neural Networks (I-PINNs)
Simulation By Deep Neural Operator (DeepONets)
PINNs Using NVIDIA Modulus
NVIDIA Modulus: Advanced Topics
Machine Learning For Engineering : A-Z

[@friedrichwilhelmhufnagel3577]

Thank you. You are great in preparing one with the apt intuition building being the first stage of understanding the actual math - thanks to the intuition building only a question of "sooner or later" any more. Great talker.

[@Dr.Mohammad_Samara]

Thank you for your comment, ideas are simple only have some layers that made it seem complex.

[@InoceramusGigas]

Great presentation!
But i think for viewers it is worth noting that one of the most powerful aspects of the FNO is its ability to learn many timesteps.
For example one can provide a single input (initial condition) and have the network output a tensor representing all time steps t0 -> t_end
As such it would require that you have representative data available for training, whether it be numerically simulated or experimentally gathered.
Great talk

[@Dr.Mohammad_Samara]

Thank you very much for important insight, this video is part of our course about deep neural operator in which we do use the data to build a simulation model using Nvidia Modulus.

[@Dr.Mohammad_Samara]

Hope you can enjoy our courses in our Podia platform:
www.courses.machinedecision.com/