Differential diagnosis of pleural mesothelioma using Logic Learning Machine

BMC Bioinformatics

Table 4 LLM quality measures for the rules shown in Table 3.

		1^st Condition		2^nd Condition		3^rd Condition		4^th Condition
n	Diag.	w%	R(c)%	w%	R(c)%	w%	R(c)%	w%	R(c)%	Cov. %
1	MPM	14.5	12.6	21.4	18.5	11.1	9.61			86.5
2	MPM	5.13	3.75	30.8	22.5	14.5	10.6			73.1
3	MPM	2.56	1.72	30.8	20.7	23.9	16.1			67.3
4	MPM	59.8	32.2	4.27	2.29					53.8
5	MTX	1.87	1.05	71.0	40.1					56.5
6	MTX	7.48	3.13	0.93	0.38	35.5	14.9			41.9
7	MTX	8.41	3.38	33.6	13.6					40.3
8	MTX	12.2	4.70	11.2	4.33	4.67	1.80			38.7
9	MTX	15.0	4.09	40.2	11.0					27.4
10	BD	29.0	20.5	4.39	3.11	2.63	1.86			70.9
11	BD	18.4	10.7	0.88	0.51	4.39	2.55	3.51	2.04	58.2
12	BD	98.2	37.5							38.2
13	BD	14.9	4.87	23.7	7.74	4.39	1.43			32.7
14	BD	9.65	2.80	16.7	4.85	2.63	0.76			29.1

Diag. = Diagnosis; w% = E(r') - E(r); R(c)% = relevance%; Cov. % = Covering percent. w and R(c) are defined according to equation (4).

ISSN: 1471-2105