Talk:Physics-informed neural networks

Hi!, interesting work!

If I understood correctly, we should try to find a citation for each phrase that we write in a wiki article. Of course, this sometimes is not possible, but, for example, in the first two sections, there are some 'claims' about the performance improvement (for example) of this new technology, that I think is worth mentioning the source.

Regarding the image: very good, but my suggestion is to increase the size or remove the illegible (at least for me) text inside it.

Acronym: since the Wikipedia audience is general and to keep high readability, I would say to not use the acronym. For the general user, is better to read 'neural network' rather than 'NN', because maybe she/he is new to this field and can easily forgot about the meaning of NN and has to jump back to find it.

Thank you for your contribution! --EdoardoRamalli (talk) 14:04, 18 June 2021 (UTC)

The topic is very interesting, and in my opinion well presented. It was not easy to convey all that information in a clear and direct way, especially considering the topic itself crosses two different scientific fields, that is to say, those of computing and physics.

Wario93 (talk) 14:20, 18 June 2021 (UTC)

The article is very well written giving a comprehensive view about the topic. I have implemented some minor revisions and my only suggestion is to add some additional references. --Francesca tagliaferri polimi (talk) 14:35, 18 June 2021 (UTC)

Partial Differential Equations (PDEs) as description of Physical laws
The article claims "Most of the physical laws that govern the dynamics of a system can be described by partial differential equations". I cannot come up with a single example that is not described by a PDE. What would be an example? I would claim, that all physical laws are described by some PDE. 130.75.31.228 (talk) 14:25, 26 January 2024 (UTC)


 * Newton's laws of motion are ODE :) 176.112.113.193 (talk) 21:32, 5 February 2024 (UTC)