“…The classification errors led to a better class separation for Arrangement 2 (1, 250, and 3,000 Hz) than Arrangement 1 (5, 250, and 2,000 Hz), the latter widely used in literature for the evaluation of peripheral nervous fibers (Hedman andSullivan, 2011, Lowenstein et al, 2008;Matsutomo et al, 2005;Nishimura et al, 2004;Pimentel et al, 2006). Additionally, it was possible to verify that classification errors occur mainly at low and medium frequencies, for both arrangements, suggesting that such stimuli can activate similar fiber groups.…”