Bias detection in multiclass models alerts you of the presence of bias so you can mitigate its distorting effects in your analysis.
-
Open a model that is associated with a multiclass classification use case. Select the
variable displaying bias or click the Review Alerts button.
-
Review the information displayed on the variable card.
# Area Description 1 Disparate Impact Data reflects discriminatory practices toward a particular demographic. Removing disparate impact from your model can produce more accountable and ethical models and, therefore, predictions that are fair and equitable. 2 Proxy Variable An explanatory variable that is highly correlated to another explanatory variable in relation to the outcome variable. 3 All Classes Use the All Classes dropdown menu to filter the chart to a specific value.