Table 4.

Performance (F1-scores) of six classifiers for each combination of features.

Model	Baseline feature set	Feature set 1	Feature set 2	Feature set 3
Logistic regression	90.2	90.3	90.2	90.5
Elastic Net	90.2	90.2	90.2	90.6
LASSOa	90.3	90.1	90.3	90.5
Random forest	90.0	90.0	90.0	90.0
Ridge classifier	90.8	90.2	90.7	90.6
XGBoostb	89.9	90.2	90.1	90.2

^aLASSO: least absolute shrinkage and selection operator.

^bXGBoost: extreme gradient boosting.