Table S2.

Mean error rate (ER) and SD for various classification methods for AMA vs MMA

Classification method	Runs	Parameters	ER, %
PCA-LR	100	L=10	33.55±8.55
PCA-LDA	100	L=10	34.02±8.49
FPCA-LR	100	L=9	33.71±8.26
FPCA-LDA	100	L=11	34.29±8.62

Note: The Parameters column in Table S2 contains the information on the number of retained principal components for both PCA-LR and PCA-LDA methods at the last run and the number of functional principal components for both FPCA-LR and FPCA-LDA methods at the last run.

Abbreviations: AMA, α-mycolic acid; FPCA, functional PCA; LDA, linear discriminant analysis; LR, logistic regression; MMA, methoxy-MA; PCA, principal-component analysis.