Causal Theories: A Categorical Perspective on Bayesian Networks

In this dissertation we develop a new formal graphical framework for causal reasoning. Starting with a review of monoidal categories and their associated graphical languages, we then revisit probability theory from a categorical perspective and introduce Bayesian networks, an existing structure for describing causal relationships. Motivated by these, we propose a new algebraic structure, which we term a causal theory. These take the form of a symmetric monoidal category, with the objects representing variables and morphisms ways of deducing information about one variable from another. A major advantage of reasoning with these structures is that the resulting graphical representations of morphisms match well with intuitions for flows of information between these variables. These categories can then be modelled in other categories, providing concrete interpretations for the variables and morphisms. In particular, we shall see that models in the category of measurable spaces and stochastic maps provide a slight generalisation of Bayesian networks, and naturally form a category themselves. We conclude with a discussion of this category, classifying the morphisms and discussing some basic universal constructions. ERRATA: (i) Pages 41-42: Objects of a causal theory are words, not collections, in $V$, and we include swaps as generating morphisms, subject to the identities defining a symmetric monoidal category. (ii) Page 46: A causal model is a strong symmetric monoidal functor.

[1]  E. Stein,et al.  Real Analysis: Measure Theory, Integration, and Hilbert Spaces , 2005 .

[2]  Judea Pearl,et al.  Causal networks: semantics and expressiveness , 2013, UAI.

[3]  S. Ulam,et al.  Zur Masstheorie in der allgemeinen Mengenlehre , 1930 .

[4]  S. Lane Categories for the Working Mathematician , 1971 .

[5]  Ernst-Erich Doberkat,et al.  Kleisli morphisms and randomized congruences for the Giry monad , 2007 .

[6]  Jared Culbertson,et al.  A Categorical Foundation for Bayesian Probability , 2012, Applied Categorical Structures.

[7]  J. Pearl Causality: Models, Reasoning and Inference , 2000 .

[8]  Arnold M. Faden,et al.  The Existence of Regular Conditional Probabilities: Necessary and Sufficient Conditions , 1985 .

[9]  Jon Williamson,et al.  Bayesian Nets and Causality: Philosophical and Computational Foundations , 2005 .

[10]  Tomas Hrycej,et al.  Gibbs Sampling in Bayesian Networks , 1990, Artif. Intell..

[11]  Dusko Pavlovic,et al.  Quantum and Classical Structures in Nondeterminstic Computation , 2008, QI.

[12]  A. Joyal,et al.  The geometry of tensor calculus, I , 1991 .

[13]  Bernd Sturmfels,et al.  Algebraic geometry of Bayesian networks , 2005, J. Symb. Comput..

[14]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems , 1988 .

[15]  Steffen L. Lauritzen,et al.  Lectures on Contingency Tables , 2002 .

[16]  P. Panangaden,et al.  Nuclear and trace ideals in tensored-categories , 1998, math/9805102.

[17]  Dudley,et al.  Real Analysis and Probability: Measurability: Borel Isomorphism and Analytic Sets , 2002 .

[18]  Michael Wendt The category of disintegration , 1994 .

[19]  Samson Abramsky,et al.  A categorical semantics of quantum protocols , 2004, LICS 2004.

[20]  P. Panangaden Probabilistic Relations , 1998 .

[21]  Robert W. Spekkens,et al.  Picturing classical and quantum Bayesian inference , 2011, Synthese.

[22]  S. Maclane,et al.  Categories for the Working Mathematician , 1971 .

[23]  P. Selinger A Survey of Graphical Languages for Monoidal Categories , 2009, 0908.3347.

[24]  Michèle Giry,et al.  A categorical approach to probability theory , 1982 .