From Moore-Penrose to Markov via Gauss

Abstract

Markov categories are the central framework for categorical probability theory. Many important concepts from probability theory can be formalized in terms of Markov categories. In particular, conditional probability distributions and Bayes' theorem are captured via the notion of conditionals in a Markov category. Gaussian probability theory gives an example of a Markov category with conditionals, where the conditionals can be computed using the Moore-Penrose inverse. In this paper, we introduce the Gauss construction on a Moore-Penrose dagger additive category, producing a Markov category with conditionals. Applying the Gauss construction to the category of real matrices recaptures the Gaussian probability theory example, while applying it to the category of complex (resp. quaternionic) matrices gives us new Markov categories of proper complex (resp. quaternionic) Gaussian conditional distributions. Moreover, we also characterize all possible conditionals in the Gauss construction.