Constraining the solution to be a probability density

[ievlevpn]

Hi!

I'm solving a system of two coupled Fokker-Planck equations and so far I didn't find a way to implement the integral constraint ensuring that my solution is a pdf.

What's the "correct" way to do this? If through a loss term, how do I implement it? If using the output map, how to do this correctly?

Thanks in advance for your help!

Pavel

[lululxvi]

Check the integral PDE in the DeepXDE paper.