A Java-based tool for the design of classification microarrays

BMC Bioinformatics

Table 2 Classification accuracy of mixed-plasmid data with model-based clustering. PA is the prediction accuracy.

	Number of clusters of probes, κ
	2		5		10		20		30		40
Number of top-ranked probes	PA (%)	No. of probes	PA (%)	No. of probes	PA (%)	No. of probes	PA (%)	No. of probes	PA (%)	No. of probes	PA (%)	No. of probes
100	83.72	2	95.35	5	95.35	10	95.35	19	95.35	36	95.35	33
150	53.49	2	90.70	5	93.02	10	93.02	20	93.02	36	93.02	28
200	79.07	2	93.02	5	93.02	10	93.02	20	93.02	36	93.02	35
250	76.74	2	95.35	5	95.35	10	95.35	20	95.35	35	93.02	32
300	69.77	2	93.02	5	93.02	10	95.35	20	95.35	34	95.35	35
350	67.44	2	93.02	5	93.02	10	93.02	20	93.02	35	93.02	35
400	69.77	2	93.02	5	93.02	10	93.02	20	93.02	37	95.35	35

ISSN: 1471-2105