A genetic algorithm-Bayesian network approach for the analysis of metabolomics and spectroscopic data: application to the rapid identification of Bacillus spores and classification of Bacillus species

Table 2 Species: classification results for all bacilli (including vegetative and sporulated cases).

Confusion Matrices Species - all bacilli
		predicted class
		all 150 m / z intensities							22 GA selected m / z intensities
		sph	sub	lic	lat	cer	amy	meg	sph	sub	lic	lat	cer	amy	meg
actual class	sph sub lic lat cer amy meg	29 3 1 0 0 1 1	1 24 1 0 0 3 0	0 0 21 0 0 0 3	0 1 0 23 1 0 0	0 0 0 1 14 0 0	0 14 7 0 0 26 0	0 0 0 0 15 0 26	30 3 1 1 3 0 6	0 32 0 0 0 3 0	0 0 22 0 0 1 0	0 0 1 23 0 0 0	0 0 0 0 25 0 0	0 7 6 0 0 26 0	0 0 0 0 2 0 24
	Statistics
	class	TP	FP	precision	sensitivity				TP	FP	precision	sensitivity
		all 150 m/z intensities							22 GA selected m/z intensities
	sph	0.967	0.032	0.829	0.967				1	0.075	0.682	1
	sub	0.571	0.029	0.828	0.571				0.762	0.017	0.914	0.762
	lic	0.7	0.016	0.875	0.7				0.733	0.005	0.957	0.733
	lat	0.958	0.01	0.92	0.958				0.958	0.005	0.958	0.958
	cer	0.467	0.005	0.933	0.467				0.833	0	1	0.833
	amy	0.867	0.113	0.553	0.867				0.867	0.07	0.667	0.867
	meg	0.867	0.081	0.634	0.867				0.8	0.011	0.923	0.8
	w. avg.	0.755	0.041	0.794	0.755				0.843	0.026	0.871	0.843

ISSN: 1471-2105