Seminar, Hosted by Topos Institute Colloquium
Ergodic Theory, categorically
Abstract
Categorical treatments of famous theorems about invariant measures, including the de Finetti theorem, the strong law of large numbers and ergodic decompositions, have emerged in recent years. Tomas Gonda synthesises these developments by introducing a mixing morphism for an abstract dynamical system in a Markov category, which axiomatises finding an average of the dynamics that is itself invariant and is closely tied to the pointwise ergodic theorem. The talk explores the abstract consequences of having a mixing morphism and several concrete instances in familiar Markov categories: general probabilistic systems, linear systems with Gaussian noise, and possibilistic uncertainty.
Topics
- ergodic theory
- invariant measures, search World Wide
- de Finetti theorem, search World Wide
- strong law of large numbers, broader topic: law of large numbers
- ergodic decompositions, search World Wide
- mixing morphism, search World Wide
- Markov category, search World Wide
- pointwise ergodic theorem, search World Wide
- Gaussian noise, broader topic: noise
- possibilistic uncertainty, broader topic: uncertainty