plot.LBBNN_Net

Given a trained <code>LBBNN_Net</code> model, this function produces either:<ul>
<li>Global plot: a visualization of the network structure,
showing only the active paths.</li>
<li>Local explanation: a plot of the local
explanation for a single input sample, including error bars obtained
from Monte Carlo sampling of the network weights.</li>
</ul>

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>.

Lars Skaaret-Lund

LBBNN

Latent Binary Bayesian Neural Networks Using 'torch'

Aliaksandr Hubin

Eirik Høyheim

plot.LBBNN_Net function

<dl><dt>x</dt>
<dd>An instance of <code>LBBNN_Net</code>.</dd>
<dt>type</dt>
<dd>Either <code>"global"</code> or <code>"local"</code>.</dd>
<dt>data</dt>
<dd>If local is chosen, one sample must be provided to obtain the explanation. Must be a <code>torch::torch_tensor</code> of shape <code>(1,p)</code>.</dd>
<dt>num_samples</dt>
<dd>integer, how many samples to use for model averaging over the weights in case of local explanations.</dd>
<dt>...</dt>
<dd>further arguments passed to or from other methods.</dd></dl>

Arguments

Plot LBBNN_Net objects — plot.LBBNN_Net

<dl>

<dt>x</dt>
<dd>An instance of <code>LBBNN_Net</code>.</dd>


<dt>type</dt>
<dd>Either <code>"global"</code> or <code>"local"</code>.</dd>


<dt>data</dt>
<dd>If local is chosen, one sample must be provided to obtain the explanation. Must be a <code>torch::torch_tensor</code> of shape <code>(1,p)</code>.</dd>


<dt>num_samples</dt>
<dd>integer, how many samples to use for model averaging over the weights in case of local explanations.</dd>


<dt>...</dt>
<dd>further arguments passed to or from other methods.</dd>

</dl>

plot.LBBNN_Net: Plot `LBBNN_Net` objects

Description

Usage

Value

Arguments