Package: LBBNN
Title: Latent Binary Bayesian Neural Networks Using 'torch'
Version: 0.1.6
Authors@R: c(person("Lars", "Skaaret-Lund", email = "lars.skaaret-lund@nmbu.no", role = c("aut", "cre")), person("Aliaksandr", "Hubin", email = "aliaksandr.hubin@nmbu.no", role = c("aut")), person("Eirik", "Høyheim", email = "eirik.hoyheim@ffi.no", role = "aut"))
Maintainer: Lars Skaaret-Lund <lars.skaaret-lund@nmbu.no>
Description: Latent binary Bayesian neural networks (LBBNNs) are
        implemented using 'torch', an R interface to the LibTorch
        backend. Supports mean-field variational inference as well as
        flexible variational posteriors using normalizing flows. The
        standard LBBNN implementation follows Hubin and Storvik (2024)
        <doi:10.3390/math12060788>, using the local reparametrization
        trick as in Skaaret-Lund et al. (2024)
        <https://openreview.net/pdf?id=d6kqUKzG3V>. Input-skip
        connections are also supported, as described in Høyheim et al.
        (2025) <doi:10.48550/arXiv.2503.10496>.
License: MIT + file LICENSE
Encoding: UTF-8
RoxygenNote: 7.3.2
Language: en-US
Suggests: testthat (>= 3.0.0), knitr, rmarkdown, torchvision
Config/testthat/edition: 3
Depends: R (>= 3.5)
LazyData: true
VignetteBuilder: knitr
Imports: ggplot2, torch, igraph, coro, svglite
NeedsCompilation: no
Packaged: 2026-06-30 19:27:24 UTC; root
Author: Lars Skaaret-Lund [aut, cre], Aliaksandr Hubin [aut], Eirik
        Høyheim [aut]
Repository: https://cran.r-universe.dev
Date/Publication: 2026-06-30 14:00:02 UTC
RemoteUrl: https://github.com/cran/LBBNN
RemoteRef: HEAD
RemoteSha: f1b530927619d3118055f54196ec803501c47aeb