Adaptive Ensemble with Human Memorizing Characteristics for Data Stream Mining

Table 5

Results of average chunk testing time [10⁻⁶ second].

Dataset	Win	SEA	AWE	ACE	MAE
SEA	0.30	2.00	2.50	4.50	2.00
Tree	0.50	3.50	3.50	21.50	3.55
Hyp	1.50	3.50	4.00	20.50	3.50
RBF	0.50	3.55	3.50	33.00	3.05
LED	0.50	4.00	4.00	5.00	3.50
Adult	0.50	2.20	3.50	4.50	3.50
Bank	1.50	17.20	19.50	153.00	23.00
Conn	0.45	14.76	23.10	33.40	22.10
Cover	0.30	0.50	0.40	1.76	0.40
EGG	1.30	3.20	4.50	5.00	4.50
Elec	22.50	41.90	33.00	92.70	33.00
Page	10.50	45.00	43.00	123.00	75.00
Person	0.30	11.50	11.70	16.50	12.30
Poker	2.50	4.70	3.10	9.10	3.30
Robot	0.30	35.30	32.50	136.00	62.50
Mean	2.90	12.85	12.79	43.96	17.01