Measure Transformer Semantics for Bayesian Machine Learning

By Johannes Borgström, A. Gordon, Michael Greenberg, James Margetson, Jurgen Van Gael. 2011

109 citations recorded by the source corpus.

The Bayesian approach to machine learning amounts to computing posterior distributions of random variables from a probabilistic model of how the variables are related (that is, a prior distribution ...

Subfields: Artificial Intelligence, Probabilistic & Bayesian Methods, Program Analysis & Verification

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Probabilistic & Bayesian Methods

In Program Analysis & Verification

Citations in the corpus

PSI: Exact Symbolic Inference for Probabilistic Programs (2016, methodology): “The general approaches include sampling-based Monte Carlo methods [20,40,21,19,39,46,32] and projections to convenient probability distributions, such as variational inference [34,8] or discretization [31,12].”
Practical probabilistic programming with monads (2015, background): “Borgström et al. (2013) develop a compositional semantics of first-order functional programs, including conditioning on zero-probability events, based on measure transformers, but they do not consider the higher-order case.”
Semantics for probabilistic programming: higher-order functions, continuous distributions, and soft constraints (2016, background): “One intere s ing next step is to use these tools to study other old or new lan guage features and concepts (such as recursion, function me moisation [32], measure-zero conditioning [5], disintegration [2, 34], and exchangeability [11, 21, 37]) that have been experimented w ith in the context of prob”
Approximate Computing: A Survey (2016, methodology): “NET [24], [33] uses various approximate and exact inference engines, each of which has different restrictions.”
A Program Logic for Union Bounds (2016, unknown): “More recent works study the semantics of probabilistic programs for applications like statistical computations [9], probabilistic inference for machine learning [10], probabilistic modeling for software deﬁned networks [17], and more.”
Commutative Semantics for Probabilistic Programming (2017, background): “…Hughes and others have eschewed monads and given categorical semantics for ﬁrst order languages in terms of Freyd categories [25] or Arrows [16] (see also [2,17,41]), and the idea of structuring the ﬁnite kernels as an Arrow already appears in the work of B¨orgstrom et al. [3] (see also [36,44]).”