Efficient Model Evidence Computation in Tree-structured Factor Graphs

Nguyen, Hoang M.H.; Erp, Bart van; Senoz, Ismail; Vries, Bert de

doi:10.1109/sips55645.2022.9919250

Cited by 2 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The presented approach is especially convenient when the model allows for the use of scale factors (Ch.6, [ 38 , 39 ]). In this way, we can efficiently compute the model evidence as shown in [ 39 ].…”

Section: Discussionmentioning

confidence: 99%

“…A limitation of the scale factors is that they can only be efficiently computed when the model submits to exact inference [ 39 ]. Extensions of the scale factors towards a variational setting would allow the use of the mixture node with a bigger variety of models.…”

Section: Discussionmentioning

confidence: 99%

“…Specifically, we build on the results of the recently developed scale factors (Ch. 6, [ 38 , 39 ]), which we introduce in Section 3.3 . These scale factors support the efficient tracking of local summaries of model evidences, thus enabling model comparison in generic mixture models; see Section 4 and Section 5 .…”

Section: Related Workmentioning

confidence: 99%

“…The term

denotes the scaling of the message

, also known as the scale factor (Ch. 6, [ 38 , 39 ]). Scale factors can be interpreted as local summaries of the model evidence that are passed along the graph.…”

Section: Background Materialsmentioning

confidence: 99%

See 3 more Smart Citations

Automating Model Comparison in Factor Graphs

van Erp,

Nuijten,

van de Laar

et al. 2023

Entropy

Self Cite

View full text Add to dashboard Cite

Bayesian state and parameter estimation are automated effectively in a variety of probabilistic programming languages. The process of model comparison on the other hand, which still requires error-prone and time-consuming manual derivations, is often overlooked despite its importance. This paper efficiently automates Bayesian model averaging, selection, and combination by message passing on a Forney-style factor graph with a custom mixture node. Parameter and state inference, and model comparison can then be executed simultaneously using message passing with scale factors. This approach shortens the model design cycle and allows for the straightforward extension to hierarchical and temporal model priors to accommodate for modeling complicated time-varying processes.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

“…The term

denotes the scaling of the message

, also known as the scale factor (Ch. 6, [ 38 , 39 ]). Scale factors can be interpreted as local summaries of the model evidence that are passed along the graph.…”

Section: Background Materialsmentioning

confidence: 99%

See 2 more Smart Citations

Automating Model Comparison in Factor Graphs

van Erp,

Nuijten,

van de Laar

et al. 2023

Entropy

Self Cite

View full text Add to dashboard Cite

show abstract

“…Over the past few years, the inference methods in this package have been tested on many advanced probabilistic models, resulting in several publications in highly ranked journals such as Entropy (Podusenko, Kouw, et al, 2021;Şenöz et al, 2021), Frontiers (Podusenko, Erp, Koudahl, et al, 2021), and conferences such as MLSP-2021 (Podusenko, Erp, Bagaev, et al, 2021), EUSIPCO-2022Podusenko et al, 2022) and SiPS (Nguyen et al, 2022).…”

mentioning

confidence: 99%

RxInfer: A Julia package for reactive real-time Bayesian inference

Bagaev¹,

Podusenko²,

Vries³

2023

JOSS

Self Cite

View full text Add to dashboard Cite

Bayesian inference realizes optimal information processing through a full commitment to reasoning by probability theory. The Bayesian framework is positioned at the core of modern AI technology for applications such as speech and image recognition and generation, medical analysis, robot navigation, and more. The framework describes how a rational agent should update its beliefs when new information is revealed by the agent's environment. Unfortunately, perfect Bayesian reasoning is generally intractable, since calculations of (often) very high-dimensional integrals are required for many models of interest. As a result, a number of numerical algorithms for approximating Bayesian inference have been developed and implemented in probabilistic programming packages. Successful methods include the Laplace approximation (Gelman et al., 2015), variants of Monte Carlo (MC) sampling (Salimans et al., n.d.), Variational Inference (VI) (Blei et al., 2017), Automatic-Differentiation Variational Inference (ADVI) , and Black-Box Variational Inference (BBVI) (Bamler & Mandt, 2017).

show abstract

Efficient Model Evidence Computation in Tree-structured Factor Graphs

Cited by 2 publications

References 0 publications

Automating Model Comparison in Factor Graphs

Automating Model Comparison in Factor Graphs

RxInfer: A Julia package for reactive real-time Bayesian inference

Contact Info

Product

Resources

About