vaeac_check_epoch_values

Function that checks provided epoch arguments

internal

Complex machine learning models are often hard to interpret. However, in
many situations it is crucial to understand and explain why a model made a specific
prediction. Shapley values is the only method for such prediction explanation framework
with a solid theoretical foundation. Previously known methods for estimating the Shapley
values do, however, assume feature independence. This package implements methods which accounts for any feature
dependence, and thereby produces more accurate estimates of the true Shapley values.
An accompanying 'Python' wrapper ('shaprpy') is available through the GitHub repository.

Martin Jullum

shapr

Prediction Explanation with Dependence-Aware Shapley Values

Lars Henry Berge Olsen

Annabelle Redelmeier

Jon Lachmann

Nikolai Sellereite

Anders Løland

Jens Christian Wahl

Camilla Lingjærde

Norsk Regnesentral 

vaeac_check_epoch_values function

<dl><dt>epochs</dt>
<dd>Positive integer (default is <code>100</code>). The number of epochs to train the final vaeac model.
This includes <code>epochs_initiation_phase</code>, where the default is <code>2</code>.</dd>
<dt>epochs_initiation_phase</dt>
<dd>Positive integer (default is <code>2</code>). The number of epochs to run each of the
<code>n_vaeacs_initialize</code> <code>vaeac</code> models before continuing to train only the best performing model.</dd>
<dt>epochs_early_stopping</dt>
<dd>Positive integer (default is <code>NULL</code>). The training stops if there has been no
improvement in the validation IWAE for <code>epochs_early_stopping</code> epochs. If the user wants the training process
to be solely based on this training criterion, then <code>epochs</code> in <code>explain()</code> should be set to a large
number. If <code>NULL</code>, then <code>shapr</code> will internally set <code>epochs_early_stopping = vaeac.epochs</code> such that early
stopping does not occur.</dd>
<dt>save_every_nth_epoch</dt>
<dd>Positive integer (default is <code>NULL</code>). If provided, then the vaeac model after
every <code>save_every_nth_epoch</code>th epoch will be saved.</dd></dl>

Arguments

Author

Function that checks provided epoch arguments — vaeac_check_epoch_values

<dl>

<dt>epochs</dt>
<dd>Positive integer (default is <code>100</code>). The number of epochs to train the final vaeac model.
This includes <code>epochs_initiation_phase</code>, where the default is <code>2</code>.</dd>


<dt>epochs_initiation_phase</dt>
<dd>Positive integer (default is <code>2</code>). The number of epochs to run each of the
<code>n_vaeacs_initialize</code> <code>vaeac</code> models before continuing to train only the best performing model.</dd>


<dt>epochs_early_stopping</dt>
<dd>Positive integer (default is <code>NULL</code>). The training stops if there has been no
improvement in the validation IWAE for <code>epochs_early_stopping</code> epochs. If the user wants the training process
to be solely based on this training criterion, then <code>epochs</code> in <code>explain()</code> should be set to a large
number. If <code>NULL</code>, then <code>shapr</code> will internally set <code>epochs_early_stopping = vaeac.epochs</code> such that early
stopping does not occur.</dd>


<dt>save_every_nth_epoch</dt>
<dd>Positive integer (default is <code>NULL</code>). If provided, then the vaeac model after
every <code>save_every_nth_epoch</code>th epoch will be saved.</dd>

</dl>

Function that checks provided epoch arguments

vaeac_check_epoch_values: Function that checks provided epoch arguments

Description

Usage

Value

Arguments

Author