get_feature_specs

Get feature specifications from the model

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 

get_feature_specs function

<dl><dt>get_model_specs</dt>
<dd>Function.
An optional function for checking model/data consistency when <code>model</code> is not natively supported.
(Run <code>get_supported_models()</code> for a list of natively supported models.)
The function takes <code>model</code> as an argument and provides a list with 3 elements:<dl>
<dt>labels</dt>
<dd>Character vector with the names of each feature.</dd></dl></dd><dt>classes</dt>
<dd>Character vector with the class of each feature.</dd><dt>factor_levels</dt>
<dd>Character vector with the levels for any categorical features.</dd>
</dl>If <code>NULL</code> (the default), internal functions are used for natively supported model classes, and checking is
disabled for unsupported model classes.
Can also be used to override the default function for natively supported model classes.
<dt>model</dt>
<dd>Model object.
The model whose predictions you want to explain.
Run <code>get_supported_models()</code>
for a table of which models <code>explain</code> supports natively. Unsupported models
can still be explained by passing <code>predict_model</code> and (optionally) <code>get_model_specs</code>,
see details for more information.</dd>

Arguments

Get feature specifications from the model — get_feature_specs

<dl>

<dt>get_model_specs</dt>
<dd>Function.
An optional function for checking model/data consistency when <code>model</code> is not natively supported.
(Run <code>get_supported_models()</code> for a list of natively supported models.)
The function takes <code>model</code> as an argument and provides a list with 3 elements:<dl>
<dt>labels</dt>
<dd>Character vector with the names of each feature.</dd>

<dt>classes</dt>
<dd>Character vector with the class of each feature.</dd>

<dt>factor_levels</dt>
<dd>Character vector with the levels for any categorical features.</dd>


</dl>If <code>NULL</code> (the default), internal functions are used for natively supported model classes, and checking is
disabled for unsupported model classes.
Can also be used to override the default function for natively supported model classes.</dd>


<dt>model</dt>
<dd>Model object.
The model whose predictions you want to explain.
Run <code>get_supported_models()</code>
for a table of which models <code>explain</code> supports natively. Unsupported models
can still be explained by passing <code>predict_model</code> and (optionally) <code>get_model_specs</code>,
see details for more information.</dd>

</dl>

Get feature specifications from the model

get_feature_specs: Get feature specifications from the model

Description

Usage

Arguments