group_model_performance

Special method for model performance evaluation. Counts number of tp, tn, fp, fn for each subgroup (and therefore potentially distinct cutoff), sums afterwards.

Measure fairness metrics in one place for many models. Check how big is model's bias towards different races, sex, nationalities etc. Use measures such as Statistical Parity, Equal odds to detect the discrimination against unprivileged groups. Visualize the bias using heatmap, radar plot, biplot, bar chart (and more!). There are various pre-processing and post-processing bias mitigation algorithms implemented. Package also supports calculating fairness metrics for regression models. Find more details in (Wiśniewski, Biecek (2021)) <arXiv:2104.00507>.

Jakub Wi<c5><9b>niewski

fairmodels

Flexible Tool for Bias Detection, Visualization, and Mitigation

Jakub Wiśniewski

Przemysław Biecek

group_model_performance function

<dl><dt>x</dt>
<dd>object created with <code>explain</code></dd>
<dt>protected</dt>
<dd>factor, vector with levels as subgroups</dd>
<dt>cutoff</dt>
<dd>vector of thresholds for each subgroup</dd>
<dt>performance_metric</dt>
<dd>name of performance metric</dd></dl>

Arguments

Group model performance — group_model_performance

<dl>

<dt>x</dt>
<dd>object created with <code>explain</code></dd>


<dt>protected</dt>
<dd>factor, vector with levels as subgroups</dd>


<dt>cutoff</dt>
<dd>vector of thresholds for each subgroup</dd>


<dt>performance_metric</dt>
<dd>name of performance metric</dd>

</dl>

group_model_performance: Group model performance

Description

Usage

Value

Arguments