Stream-Based Extreme Learning Machine Approach for Big Data Problems

Table 5

Big Data results.

Key	Used Labels	Ac	AUC
EALM
Splice	422.80 ± 14.86 (42.28%)	0.83 ± 0.01	0.83 ± 0.01
IJCNN1	3244.70 ± 112.04 (6.49%)	0.96 ± 0.01	0.86 ± 0.02
Web Gaussian	2108.50 ± 82.46 (4.23%)	0.98 ± 0.01	0.77 ± 0.01
Adult	4285.70 ± 88.35 (18.88%)	0.84 ± 0.01	0.76 ± 0.01
ELM2012
Splice	1000 (100%)	0.85 ± 0.01	0.85 ± 0.01
IJCNN1	49990 (100%)	0.95 ± 0.01	0.79 ± 0.01
Web Gaussian	49749 (100%)	0.98 ± 0.01	0.79 ± 0.01
Adult	22696 (100%)	0.84 ± 0.00	0.74 ± 0.00
SVM
Splice	1000 (100%)	0.81 ± 0.02	0.81 ± 0.02
IJCNN1	49990 (100%)	0.97 ± 0.01	0.91 ± 0.01
Web Gaussian	49749 (100%)	0.99 ± 0.01	0.87 ± 0.01
Adult	22696 (100%)	0.84 ± 0.01	0.75 ± 0.01