zero_one

Calculates the per-observation 0/1 loss as $$
 t_i \neq r_1.
$$
Measure to compare true observed labels with predicted
labels
in multiclass classification tasks.
Note that this is an unaggregated measure, returning the losses per observation.

Implements multiple performance measures for
supervised learning.  Includes over 40 measures for regression and
classification. Additionally, meta information about the performance
measures can be queried, e.g. what the best and worst possible
performances scores are.

Michel Lang

mlr3measures

Performance Measures for 'mlr3'

Martin Binder

zero_one function

<dl><dt>truth</dt>
<dd>(<code><a href="/link/factor()?package=mlr3measures&version=0.5.0" data-mini-rdoc="mlr3measures::factor()">factor()</a></code>) 
True (observed) labels.
Must have the same levels and length as <code>response</code>.</dd>
<dt>response</dt>
<dd>(<code><a href="/link/factor()?package=mlr3measures&version=0.5.0" data-mini-rdoc="mlr3measures::factor()">factor()</a></code>) 
Predicted response labels.
Must have the same levels and length as <code>truth</code>.</dd>
<dt>...</dt>
<dd>(<code>any</code>) 
Additional arguments. Currently ignored.</dd></dl>

Arguments

<ul>
<li>Type: <code>"classif"</code></li>
<li>Range (per observation): $[0, 1]$</li>
<li>Minimize (per observation): <code>TRUE</code></li>
<li>Required prediction: <code>response</code></li>
</ul>

Meta Information

Zero-One Classification Loss (per observation) — zero_one

<dl>

<dt>truth</dt>
<dd>(<code><a href='https://rdrr.io/r/base/factor.html'>factor()</a></code>) 
True (observed) labels.
Must have the same levels and length as <code>response</code>.</dd>


<dt>response</dt>
<dd>(<code><a href='https://rdrr.io/r/base/factor.html'>factor()</a></code>) 
Predicted response labels.
Must have the same levels and length as <code>truth</code>.</dd>


<dt>...</dt>
<dd>(<code>any</code>) 
Additional arguments. Currently ignored.</dd>

</dl>

<ul>
<li>Type: <code>"classif"</code></li>
<li>Range (per observation): $[0, 1]$</li>
<li>Minimize (per observation): <code>TRUE</code></li>
<li>Required prediction: <code>response</code></li>
</ul>

zero_one: Zero-One Classification Loss (per observation)

Description

Usage

Value

Arguments

Meta Information

See Also