Picturing classical and quantum Bayesian inference

Abstract

We introduce a graphical framework for Bayesian inference that is sufficiently general to accommodate not just the standard case but also recent proposals for a theory of quantum Bayesian inference wherein one considers density operators rather than probability distributions as representative of degrees of belief. The diagrammatic framework is stated in the graphical language of symmetric monoidal categories and of compact structures and Frobenius structures therein, in which Bayesian inversion boils down to transposition with respect to an appropriate compact structure. We characterize classical Bayesian inference in terms of a graphical property and demonstrate that our approach eliminates some purely conventional elements that appear in common representations thereof, such as whether degrees of belief are represented by probabilities or entropic quantities. We also introduce a quantum-like calculus wherein the Frobenius structure is noncommutative and show that it can accommodate Leifer's calculus of `conditional density operators'. The notion of conditional independence is also generalized to our graphical setting and we make some preliminary connections to the theory of Bayesian networks. Finally, we demonstrate how to construct a graphical Bayesian calculus within any dagger compact category.

38 pages, lots of pictures

References in corpus (9)

• Structure of states which satisfy strong subadditivity of quantum entropy with equality
• Quantum-Bayesian Coherence: The No-Nonsense Version
• Quantum Graphical Models and Belief Propagation
• Quantum Dynamics as an analog of Conditional Probability
• Cloning and Broadcasting in Generic Probabilistic Theories
• Finite dimensional Hilbert spaces are complete for dagger compact closed categories
• Quantum-Bayesian Coherence
• POVMs and Naimark's theorem without sums
• Ordered linear spaces and categories as frameworks for information-processing characterizations of quantum and classical theory

Cited by in corpus (27)

• Towards a Formulation of Quantum Theory as a Causally Neutral Theory of Bayesian Inference
• A synthetic approach to Markov kernels, conditional independence and theorems on sufficient statistics
• Weak Values are Interference Phenomena
• Quantum Theory and Determinism
• From time-reversal symmetry to quantum Bayes' rules
• Quantum speedup in the identification of cause-effect relations
• DisCoPy: Monoidal Categories in Python
• Quantum instruments as a foundation for both states and observables
• Von Neumann Was Not a Quantum Bayesian
• A non-commutative Bayes' theorem
• Foundations for Near-Term Quantum Natural Language Processing
• State retrieval beyond Bayes' retrodiction
• Observational entropy, coarse quantum states, and Petz recovery: information-theoretic properties and bounds
• Free gs-monoidal categories and free Markov categories
• Time-asymmetry of probabilities versus relativistic causal structure: an arrow of time
• Meaning updating of density matrices
• Inverses, disintegrations, and Bayesian inversion in quantum Markov categories
• Category Theory in Machine Learning
• Towards a Categorical Account of Conditional Probability
• Compositional Quantum Logic
• Conditional Distributions for Quantum Systems
• Lower and Upper Conditioning in Quantum Bayesian Theory
• Hidden Markov Models and the Bayes Filter in Categorical Probability
• Bipartite quantum states admitting a causal explanation
• Causality and deceit: Do androids watch action movies?
• Identification of Causal Influences in Quantum Processes
• Fast tests for probing the causal structure of quantum processes