Table 1. Performance of previous works.

Dataset	Classification system	Accuracy	Reference
BBB	Artificial Neural Networks	75.7%	[11]
	Random Forest	80.9%	[12]
	Support Vector Machines	81.5%	[11]
P-gp	Four-point Pharmacophore	62.7%	[2]
	Support Vector Machines	79.4%	[21]
	Random Forest	80.6%	[12]
estrogen	Decision Forest	80%	[13]
	Random Forest	82.8%	[12]

This table summarizes the performance of several classification methods used on the BBB, P-gp and estrogen problems.