Steve Brunton - Machine Learning for Scientific Discovery, with Examples in Fluid Mechanics

[@otter662]

brunton's online videos, lectures, learning material are enormously helpful , thank you.

[@stayinthepursuit8427]

This guy is legend ofcourse

[@zijingding4135]

I have a question for the Lorentz system: why not include d^2x/dt^2 terms in machine learning??

[@blackmail1807]

They’re redundant. Any second order equation can be written as two first order equations by introducing a new variable y=dx/dt.

[@ravikiran4495]

In matrix form (state space form) you often define the system such that you try to adjust the system in a square form,in which the considered variable of interest might be a rate or a gradient(vectors and their components), where you can further break it down using several approaches with relatively lower complexity but ofc things such as how much of coupling is involved(between the variables) can then further complicate the task depending on how "non-linear" the interaction seems to be,but we can kind of approximate this non linearity around some points if some conditions are met.

[@NoNTr1v1aL]

Absolutely amazing video! Subscribe to his KZbin channel. It has a lot of great playlists.

[@iheavense]

❤

[@hiedtchannel]

Isn’t this Steve Brunton teaching control theory?

[@KnowL-oo5po]

we need to merge phycology ,philosophy ,neuroscience, biology and physics to make an A.G.I