Table 3.

Three different classes of rules generated from dataset L2

Consequent	Antecedent	Support (%)	Confidence (%)
ALL	32847_at > 147.0	30.56	100
	36239_at > 2201.0	27.78	100
AML	39318_at < 1063.0 & 32579_at < 2285.0	34.72	100
	1325_at < 1501.5, 39318_at < 1063.0 & 32579_at < 2285.0	34.72	100
	1325_at < 1501.5, 36239_at < 214.0 & 40191_s_at < 508.5	33.33	100
	36239_at < 214.0 & 40191_s_at < 508.5	33.33	100
	39318_at < 1063.0 & 35164_at < -794.5	31.94	100
	40191_s_at < 519.0 & 36239_at < 167.0	31.94	100
	1325_at < 1501.5, 39318_at < 1063.0 & 35164_at < -794.5	31.94	100
	1325_at < 1501.5, 40191_s_at < 519.0 & 36239_at < 167.0	31.94	100
	1325_at < 1501.5, 36239_at < 214.0 & 37539_at < -362.0	31.94	100
	36239_at < 214.0 & 37539_at < -362.0	31.94	100
	37539_at < -725.5	29.17	100
	32579_at < 2285.0	36.11	96
	1325_at < 1501.5 & 32579_at < 2285.0	36.11	96
	36239_at < 214.0	40.28	93
MLL	1325_at < 201.0, 35260_at > 794.5 & 40191_s_at > 1107.5	19.44	100
	1325_at < 201.0 & 36239_at > 214.0	23.61	94
	1325_at < 201.0	37.50	67
Mean		32.02	97