A comprehensive comparison of random forests and support vector machines for microarray-based cancer classification

Table 3 Number of genes selected for each microarray dataset and gene selection method.

*Task & dataset*	*No gene selection*	*RFE*	*RFVS1*	*RFVS2*	KW	*S2N*
Dx-Alizadeh	4026	12	62	73	19	15
Dx-Alon	2000	105	16	3	15	13
Dx-Armstrong	11225	74	709	57	106	48
Dx-Bhattacharjee	12600	289	27	15	1864	653
Dx-Golub	5327	12	456	336	42	4
Dx-Khan	2308	28	17	18	15	11
Dx-Nutt	10367	1598	126	101	476	926
Dx-Pomeroy	5920	186	34	16	70	435
Dx-Ramaswamy	15009	3346	966	411	8248	10277
Dx-Ramaswamy2	13247	1576	12	4	4129	1364
Dx-Shipp	5469	8	15	6	13	89
Dx-Singh	10509	157	58	21	22	38
Dx-Staunton	5726	169	152	73	93	97
Dx-Su	12533	2429	845	320	1318	1927
Px-Beer	7129	201	15	7	953	1380
Px-Bhattacharjee	12600	21	46	7	138	61
Px-Iizuka	7070	103	38	7	168	185
Px-Pomeroy	7129	70	29	13	445	439
Px-Rosenwald	7399	2338	124	27	3201	3897
Px-Veer	24188	1056	124	20	5388	4405
Px-Veer2	24188	491	149	39	1194	1764
Px-Yeoh	12240	1187	21	6	3077	1869

ISSN: 1471-2105