Fig. 5. Decoding performance to predict a given stimulus using sensory and or resulted choice using decoding signals. (A) Decoding performance of SVM classifier using sensory signals in the sample period (-2.6~-1.3 s) to predict the given stimulus across whole time points in the tactile delayed-response task. Shaded regions indicate standard deviation across iterations. Each dot indicates the time point where decoding accuracy is significantly different (purple dot: hit vs. the error trials, yellow dot: hit vs. chance level, green dot: error vs. chance level) (Bonferroni post hoc analysis, p<0.001). (B) Decoding performance of SVM classifier using choice signals in the delay period (-1.3~0 s) to predict the resulted lick direction across whole time points in the tactile delayed-response task. Shaded regions indicate standard deviation across iterations. Each dot indicates the time point where decoding accuracy is significantly different (purple dot: hit vs. the error trials, yellow dot: hit vs. chance level, green dot: error vs. chance level) (Bonferroni post hoc analysis, p<0.001). Similarly, (C) decoding accuracy of SVM classifier using sensory signals and (D) choice signals in the auditory delayed-response task.
© Exp Neurobiol
