Stream-Based Extreme Learning Machine Approach for Big Data Problems

Table 3

Comparison of the hidden layer size.

EALM
Key	100 neurons		500 neurons		1000 neurons
	AL Labels	AUC	AL Labels	AUC	AL Labels	AUC
HRT	76.92 ± 3.36	0.82 ± 0.02	80.09 ± 1.58	0.83 ± 0.01	78.17 ± 1.09	0.83 ± 0.01
WBCO	36.47 ± 2.82	0.97 ± 0.01	36.98 ± 0.93	0.97 ± 0.00	37.16 ± 0.70	0.97 ± 0.00
WBCD	81.28 ± 5.36	0.96 ± 0.01	76.71 ± 1.01	0.97 ± 0.00	73.65 ± 0.97	0.97 ± 0.00
PIMA	177.63 ± 8.64	0.72 ± 0.01	172.96 ± 3.42	0.71 ± 0.01	168.69 ± 2.38	0.72 ± 0.01
SNR	95.97 ± 4.19	0.75 ± 0.03	100.00 ± 1.34	0.78 ± 0.01	98.42 ± 1.74	0.75 ± 0.01
ION	107.83 ± 4.44	0.82 ± 0.02	102.05 ± 1.04	0.85 ± 0.01	103.59 ± 1.37	0.84 ± 0.01
AUST	122.04 ± 5.80	0.86 ± 0.01	119.59 ± 2.30	0.86 ± 0.01	124.88 ± 1.88	0.86 ± 0.01
LIV	109.73 ± 6.23	0.60 ± 0.03	107.20 ± 4.14	0.58 ± 0.03	109.84 ± 7.10	0.59 ± 0.04
GER	291.21 ± 4.64	0.63 ± 0.01	302.07 ± 3.09	0.67 ± 0.01	315.55 ± 3.11	0.68 ± 0.01
SPAM	675.86 ± 4.33	0.89 ± 0.00	597.17 ± 4.89	0.92 ± 0.00	557.18 ± 4.68	0.92 ± 0.00
	2500 neurons		5000 neurons		10000 neurons
HRT	76.94 ± 1.96	0.83 ± 0.01	76.53 ± 1.75	0.83 ± 0.01	76.34 ± 1.11	0.84 ± 0.01
WBCO	36.90 ± 0.50	0.97 ± 0.00	36.70 ± 1.07	0.97 ± 0.00	36.12 ± 0.87	0.97 ± 0.00
WBCD	75.44 ± 1.14	0.97 ± 0.00	75.35 ± 1.15	0.97 ± 0.00	75.76 ± 0.95	0.97 ± 0.01
PIMA	168.63 ± 2.73	0.72 ± 0.01	171.63 ± 3.59	0.73 ± 0.01	172.43 ± 2.83	0.71 ± 0.01
SNR	97.54 ± 1.53	0.78 ± 0.02	98.08 ± 1.49	0.78 ± 0.02	97.75 ± 2.40	0.77 ± 0.02
ION	101.56 ± 1.65	0.85 ± 0.01	101.72 ± 1.67	0.85 ± 0.01	102.97 ± 0.78	0.85 ± 0.01
AUST	119.91 ± 4.19	0.86 ± 0.01	120.75 ± 3.46	0.86 ± 0.00	121.18 ± 3.17	0.86 ± 0.01
LIV	109.22 ± 6.94	0.59 ± 0.03	106.53 ± 5.12	0.58 ± 0.03	105.84 ± 5.22	0.58 ± 0.02
GER	313.04 ± 3.09	0.68 ± 0.01	313.45 ± 3.11	0.68 ± 0.02	314.67 ± 2.90	0.68 ± 0.01
SPAM	559.81 ± 3.67	0.92 ± 0.00	561.52 ± 6.31	0.92 ± 0.00	549.60 ± 2.95	0.92 ± 0.00