An Efficient Ensemble Learning Method for Gene Microarray Classification

Table 4

Mean classification accuracy of each classification method against 8 different gene datasets.

Dataset	ICA-based RotBoost	Single Tree	Rotation Forest	AdaBoost	Bagging	SVMs
Colon	96.10 ± 0.59	93.80 ± 0.82^•	95.21 ± 0.43^•	94.97 ± 0.63^•	94.92 ± 0.50^•	96.13 ± 0.12
CNS	95.00 ± 0.28	89.92 ± 0.61^•	92.37 ± 0.83^•	95.09 ± 0.64	93.50 ± 0.79^•	93.34 ± 0.10^•
Leukemia	98.77 ± 0.03	96.60 ± 00.46^•	97.97 ± 0.38^•	98.22 ± 0.55^•	97.47 ± 0.51^•	95.64 ± 0.49^•
Breast	97.88 ± 0.45	88.50 ± 0.72^•	98.60 ± 0.63^o	98.89 ± 0.47^o	92.74 ± 0.45^•	96.84 ± 0.02^•
Lung	99.54 ± 0.11	94.36 ± 0.42^•	97.56 ± 0.23^•	96.30 ± 0.39^•	97.08 ± 0.37^•	95.56 ± 0.55^•
Ovarian	99.40 ± 0.26	99.37 ± 0.12	99.77 ± 0.07^o	99.57 ± 0.11	99.76 ± 0.08^o	98.66 ± 0.35^•
MLL	99.31 ± 0.55	96.03 ± 0.59^•	97.61 ± 0.31^•	97.63 ± 0.45^•	97.11 ± 0.55^•	96.80 ± 0.31^•
SRBCT	99.59 ± 0.16	93.96 ± 0.59^•	97.44 ± 0.41^•	98.16 ± 0.39^•	96.46 ± 0.58^•	97.23 ± 0.44^•
Win Tie Loss		7/1/0	6/0/2	5/2/1	7/0/1	7/1/0

•Specifies that RotBoost is significantly better, and ^opoints out that RotBoost is notably worse at the significance level α = 0.05.