Scientific Reports

Table 1 Quality parameters calculated for the test set using different supervised classification algorithms to distinguish women with endometriosis and the healthy control women.

From: Spectrochemical differentiation in endometriosis based on infrared spectroscopy advanced data fusion and multivariate analysis

Model	ATR-FTIR					NIR
Model	AC¹	SENS²	SPEC³	F-Score	G-Score	AC¹	SENS²	SPEC³	F-Score	G-Score
PCA⁴-LDA⁷	65	90	42	57	62	100	100	100	100	100
PCA⁴-QDA⁸	63	38	86	53	58	100	100	100	100	100
PCA⁴-SVM⁹	77	83	70	84	76	100	100	100	100	100
SPA⁵-LDA⁷	63	83	42	56	59	77	83	70	76	76
SPA⁵-QDA⁸	63	90	38	53	58	60	75	40	52	55
SPA⁵-SVM⁹	62	90	33	48	54	57	33	70	45	48
GA⁶-LDA⁷	70	70	58	63	63	100	100	100	100	100
GA⁶-QDA⁸	56	83	20	32	40	100	100	100	100	100
GA⁶-SVM⁹	73	92	60	72	74	100	100	100	100	100

¹AC, accuracy in %
²SENS, sensitivity in %
³SPEC, specificity in %
⁴PCA, principal component analysis.
⁵SPA, successive projections algorithm.
⁶GA, genetic algorithm.
⁷LDA, linear discriminant analysis.
⁸QDA, quadratic discriminant analysis.
⁹SVM, support vector machine.

Back to article page

Search

Advanced search

Quick links