Extended Variational Message Passing for Automated Approximate Bayesian Inference

Autor: Semih Akbayrak, Ivan Bocharov, Bert de Vries

Publikováno v: Entropy, Vol 23, Iss 7, p 815 (2021)

Variational Message Passing (VMP) provides an automatable and efficient algorithmic framework for approximating Bayesian inference in factorized probabilistic models that consist of conjugate exponential family distributions. The automation of Bayesi

Externí odkaz: https://doaj.org/article/9c31a773ddde4ca7859d81135f6607fb

Message Passing-based System Identification for NARMAX Models

Autor: Albert Podusenko, Semih Akbayrak, Ismail Senoz, Maarten Schoukens, Wouter M. Kouw

Publikováno v: 2022 IEEE 61st Conference on Decision and Control (CDC), 7309-7314
STARTPAGE=7309;ENDPAGE=7314;TITLE=2022 IEEE 61st Conference on Decision and Control (CDC)

We present a variational Bayesian identification procedure for polynomial NARMAX models based on message passing on a factor graph. Message passing allows us to obtain full posterior distributions for regression coefficients, precision parameters and

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::ce1c29cfa5105c9a09352ec3be980ecc
https://doi.org/10.1109/cdc51059.2022.9992891

The Switching Hierarchical Gaussian Filter

Autor: Semih Akbayrak, Albert Podusenko, Ismail Senoz, Bert de Vries, Christoph Mathys

Publikováno v: ISIT
Senoz, I, Podusenko, A, Akbayrak, S, Mathys, C & de Vries, B 2021, The switching Hierarchical Gaussian Filter . in 2021 IEEE International Symposium on Information Theory (ISIT) . IEEE, pp. 1373-1378, IEEE International Symposium on Information Theory (ISIT), Melbourne, Australia, 12/07/2021 . https://doi.org/10.1109/ISIT45174.2021.9518229
2021 IEEE International Symposium on Information Theory, ISIT 2021-Proceedings, 1373-1378
STARTPAGE=1373;ENDPAGE=1378;TITLE=2021 IEEE International Symposium on Information Theory, ISIT 2021-Proceedings

In this paper we discuss variational message passing-based (VMP) inference in a switching Hierarchical Gaussian Filter (HGF). An HGF is a flexible hierarchical state space model that supports closed-form VMP-based approximate inference for tracking o

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::c54259350b1cd967e06ce7d09a92a89d
https://doi.org/10.1109/isit45174.2021.9518229

Sum Conditioned Poisson Factorization

Autor: Ali Taylan Cemgil, Gökhan Çapan, Taha Yusuf Ceritli, Semih Akbayrak

Publikováno v: Latent Variable Analysis and Signal Separation ISBN: 9783319937632
LVA/ICA

We develop an extension to Poisson factorization, to model Multinomial data using a moment parametrization. Our construction is an alternative to the canonical construction of generalized linear models. This is achieved by defining K component Poisso

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_________::3c594e7e57fa5a1767205d9ed7abbb16
https://doi.org/10.1007/978-3-319-93764-9_3