Back to Search Start Over

Approximate Inference by Kullback-Leibler Tensor Belief Propagation

Authors :
Bert de Vries
Henk Corporaal
Sander Stuijk
Patrick W.A. Wijnings
Electronic Systems
Bayesian Intelligent Autonomous Systems
Signal Processing Systems
EAISI High Tech Systems
EAISI Foundational
Source :
2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 5850-5854, STARTPAGE=5850;ENDPAGE=5854;TITLE=2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), ICASSP
Publication Year :
2020
Publisher :
Institute of Electrical and Electronics Engineers, 2020.

Abstract

Probabilistic programming provides a structured approach to signal processing algorithm design. The design task is formulated as a generative model, and the algorithm is derived through automatic inference. Efficient inference is a major challenge; e.g., the Shafer-Shenoy algorithm (SS) performs badly on models with large treewidth, which arise from various real-world problems. We focus on reducing the size of discrete models with large treewidth by storing intermediate factors in compressed form, thereby decoupling the variables through conditioning on introduced weights. This work proposes pruning of these weights using Kullback-Leibler divergence. We adapt a strategy from the Gaussian mixture reduction literature, leading to Kullback-Leibler Tensor Belief Propagation (KL-TBP), in which we use agglomerative hierarchical clustering to subsequently merge pairs of weights. Experiments using benchmark problems show KL-TBP consistently achieves lower approximation error than existing methods with competitive runtime.

Details

Language :
English
Database :
OpenAIRE
Journal :
2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 5850-5854, STARTPAGE=5850;ENDPAGE=5854;TITLE=2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), ICASSP
Accession number :
edsair.doi.dedup.....01eaef9f2117be775d445750c2b31fc6