Fig. 2From: Prediction of autism spectrum disorder diagnosis using nonlinear measures of language-related EEG at 6 and 12 monthsInformation about features most correlated with autism diagnostic outcome for nearly overlapping 6- and 12-month analyses (n = 54). The bottom row visualizes the values for the 12-month dataset (middle row) minus the 6-month dataset (top row). A, D, G Average number of features selected from each channel. Color indicates number of features selected from a given channel. B, E, H Average count of each EEG measure across iterations (orange) and percentage of iterations that each measure was selected at least once (blue). C, F, I Average count of each wavelet across iterations (orange) and percentage of iterations that each wavelet was selected at least once (blue)Back to article page