Permutation analysis showed that the ANOVA-PLS was highly significant and devoid of overfitting (Additional file 6 ), since the permutation distribution median corresponded exactly to the expected value of 91 misclassifications for 182 randomly permuted observations in a two-class problem.
← all excerpts
The <sup>1</sup>H NMR serum metabolomics response to a two meal challenge: a cross-over dietary intervention study in healthy human volunteers.
1
—
—