Table 2 Datasets used during our experiments
From: FairSwiRL: fair semi-supervised classification with representation learning
Dataset | Instances | Original features | Post- processed features | Sensitive attribute | Target variable |
|---|---|---|---|---|---|
SYNTHETIC | – | 5 | 5 | S | Y |
ADULT | 48–842 | 14 | 107 | Sex | Income |
BANK | 45–211 | 17 | 61 | Previous campaign | Subscription |
CARD | 30–000 | 23 | 23 | Education | Default |
COMPAS | 6–127 | 53 | 18 | Ethnicity | Criminal recidivism |