Table 2.

Model parameters for best performing random forest models.

Best models-random forest
Parameter	VRE	CRE	MRSA	MDRO
cv=StratifiedKFold	n_splits=10	n_splits=10	n_splits=10	n_splits=10
estimator=RandomForestClassifier	Yes	Yes	Yes	Yes
bootstrap	True	True	True	True
max_depth	None	None	None	None
max_leaf_nodes	None	None	None	None
min_impurity_decrease	0	0	0	0
init_min_samples_leaf	1	1	1	1
init_min_samples_split	2	2	2	2
n_estimators	200	200	200	200
n_jobs	4	4	4	4
param_grid={'min_samples_leaf'}	[5, 10,..., 250]	[5, 10,..., 250]	[5, 10,..., 250]	[5, 10,..., 250]
param_grid={pre_dispatch}	2*n_jobs	2*n_jobs	2*n_jobs	2*n_jobs
param_grid={scoring}	roc_auc	roc_auc	roc_auc	roc_auc
optimal_min_samples_leaf	5	30	10	5
Threshold Bound	0.20	0.05	0.30	0.40