Knowledge Mining from Clinical Datasets Using Rough Sets and Backpropagation Neural Network

Table 7

Reduct size for (a) hepatitis, (b) breast cancer, and (c) heart disease dataset.

(a) Reduct for hepatitis dataset

Number of attributes	9	10	11	12	13	14	15	16	17
Number of subsets	48620	43758	31824	18564	8568	3060	816	153	18
Number of reducts	2	19	71	137	142	104	43	43	1

(b) Reduct for breast cancer dataset

Number of attributes	1	2	3	4	5	6	7	8
Number of subsets	9	36	84	126	126	84	36	9
Selected reducts	—	—	—	—	—	—	5	6

(c) Reduct for heart disease dataset

Number of attributes	3	4	5	6	7	8	9	10	11	12
Number of subsets	286	715	1287	1716	1716	1287	715	286	78	13
Number of reducts	9	90	365	811	1127	1043	657	280	78	13