Dynamical systems
2016-04-26 — 2016-07-27
Wherein the Linear Assumption Is Relaxed and Systems Are Considered Whose State Is a Measure or Symbol, Randomness and Chaos Are Invoked, and Ergodic Questions and Embedding Issues Are Sketched.
On this page
Remember linear time invariant systems, as made famous by signal processing? Now relax the assumption that the model is linear, or even that its state space is in \(\mathbb{R}^n\). Maybe its state is a measure, or a symbol, or whatever. Now say the word “chaos!” Pronounce the exclamation mark. Maybe it’s a random system, a stochastic process, or a deterministic process representing the evolution of the measure of a stochastic process or whatever.
(Regarding that, one day I should try to understand how Talagrand uses isoperimetric inequalities to derive concentration inequalities.)
Topics that I should connect to this one: the weird end: “nonlinear time series wizardry”, Also “sync”. And “ergodic theory”.
To wish I understood: Takens embedding, and whether it is any statistical use at all.
There is too much to do here, and it’s done better elsewhere. Therefore: Idiosyncratic notes only.
