Table 7.:

Perfomance for hdPS algorithms and Super Learners

Data Set	Method	Negative Log Likelihood	AUC	Negative Log Likelihood (Train)	AUC (Train)	Processing Time (Seconds)
NOAH	k=50, n=200	0.50	0.80	0.51	0.79	19.77
	k=100, n=200	0.50	0.80	0.50	0.80	20.69
	k=200, n=200	0.49	0.80	0.49	0.81	22.02
	k=350, n=200	0.49	0.82	0.47	0.83	25.38
	k=500, n=200	0.49	0.82	0.46	0.84	27.35
	k=750, n=500	0.50	0.81	0.45	0.85	50.58
	k=1000, n=500	0.52	0.80	0.43	0.86	57.08
	sl_baseline	0.53	0.77	0.53	0.77	1035.43
	sl_hdps	0.48	0.82	0.47	0.83	1636.48
NSAID	k=50, n=200	0.60	0.68	0.61	0.67	43.15
	k=100, n=200	0.60	0.69	0.60	0.69	43.48
	k=200, n=200	0.59	0.70	0.60	0.69	47.08
	k=350, n=200	0.60	0.69	0.59	0.70	52.99
	k=500, n=200	0.60	0.69	0.59	0.71	58.90
	k=750, n=500	0.60	0.69	0.58	0.71	112.44
	k=1000, n=500	0.61	0.69	0.58	0.72	119.28
	sl_baseline	0.61	0.67	0.61	0.66	1101.84
	sl_hdps	0.59	0.70	0.59	0.71	2075.05
VYTORIN	k=50, n=200	0.44	0.64	0.43	0.64	113.45
	k=100, n=200	0.43	0.65	0.43	0.65	116.73
	k=200, n=200	0.43	0.65	0.43	0.66	146.81
	k=350, n=200	0.43	0.65	0.42	0.67	166.18
	k=500, n=200	0.43	0.65	0.42	0.67	189.18
	k=750, n=500	0.43	0.65	0.42	0.68	315.22
	k=1000, n=500	0.43	0.65	0.42	0.68	350.45
	sl_baseline	0.42	0.69	0.42	0.70	9165.93
	sl_hdps	0.42	0.70	0.41	0.71	15743.89