Journal of Big Data

Table 15 Results of RF with CoClust for MIMIC-III

From: An enhanced random forest approach using CoClust clustering: MIMIC-III and SMS spam collection application

		Accuracy	OOB error rate (%)	AUROC
Clayton Copula	Cluster 1	0.8453	15.47	0.911
	Cluster 2	0.8667	13.33	0.949
	Cluster 3	0.9343	6.57	0.987
	Cluster 4	0.8355	16.45	0.915
	Cluster 5	0.7963	20.37	0.883
	Cluster 6	0.8254	17.46	0.894
	Cluster 7	0.8269	17.31	0.895
Frank Copula	Cluster 1	0.8453	15.47	0.911
	Cluster 2	0.8355	16.45	0.915
	Cluster 3	0.7940	20.60	0.900
	Cluster 4	0.7963	20.37	0.883
	Cluster 5	0.8254	17.46	0.894
	Cluster 6	0.8269	17.31	0.895
	Cluster 7	0.9436	5.64	0.992

Back to article page