Block diagram of the proposed classification system for ENG signals. Training is performed on the first and testing on the second half of the data. (A) Training procedure consisting of: Filtering (in order to eliminate the EMG low band, and amplifiers high band noise); Feature extraction from running observation window (ROW); Training of the classifier with stimuli type knowledge (VF, Proprioceptive, Rest as labeled); recorded during the experimentation. (B) Test procedure: filtering and feature extraction are first steps (both the same as during the training); then the classifiers (trained in A) answer is post-processed by means of majority vote rule. Evaluation is carried out by report between correctly classified instances and all samples in each test set.