Table 5.

Sample classification after LDA to discriminate between CSF samples from patients with Relapse–remitting multiple sclerosis (RRMS) vs. other diseases (inflammatory and non-inflammatory) of the CNS (INF and N-INF) after data reduction with PCA.

	Disease	Predicted Group Membershipb.c		Total
		RRMS	INF and N-INF
Original
Count	RRMS	55	14	69
	INF and N-INF	11	58	69
%	RRMS	79.7	20.3	100
	INF and N-INF	15.9	84.1	100
Cross-validateda
Count	RRMS	55	14	69
	INF and N-INF	11	58	69
%	RRMS	79.7	20.3	100
	INF and N-INF	15.9	84.1	100

^aCross-validation is done only for those sets of samples analyzed. In cross-validation, each sample is classified by the functions derived from all samples other the one under analysis

^b81.9% of original grouped samples correctly classified

^c81.9% of cross-validated grouped samples correctly classified