The main ingredient in proving this is the observation that the coordinate-change maps must be linear on infinitesimals, so the infinitesimal neighborhood of a point is a vector space in a canonical way, which is isomorphic to the tangent space at that point.
I.e it's just a *discrete* dynamical system with the constraint that each step is infinitesimal.
The idea in this one, that you can classify the behavior around equilibria using linear algebra (because the dynamics are "locally approximately linear") is very cool.
https://www.youtube.com/watch?v=7Ewe_tVa5Fs&list=PLUeHTafWecAUqSh3Gy0NNr7H3OsXoC-aK&index=21
Watching these videos now, they're really good 😀
---
RT @sarah_zrf
im gonna do it. im gonna learn dynamics. im gonna watch the videos https://twitter.com/RossDynamicsLab/status/1432690088723587076
https://twitter.com/sarah_zrf/status/1433074184578600961
If the sausages in the first picture had been dyed red that one alone would've pretty much nailed it down tbh.
Using only food, where did you grow up?
---
RT @mechanical_monk
god bless the motherland https://twitter.com/made_in_cosmos/status/1435181546198155266 https://t.co/2yhJhHlv37
https://twitter.com/mechanical_monk/status/1435238587369500678
I also think nobody (probably not even economists who used ergodicity as a technical assumption) ever believed that "this strategy has a high expected payoff" meant you were guaranteed a high payoff eventually - that would be a weird way to talk about probability!
Applied algebraic abstractologist. Trying to get the heavens into my head.
"Elsk – og berik med drøm – alt stort som var!
Gå mot det ukjente, fravrist det svar!"