Automated EEG sentence classification using novel dynamic-sized binary pattern and multilevel discrete wavelet transform techniques with TSEEG database

•Two new EEG sentence classification datasets were collected.•A new generation local feature extraction function was proposed.•A new unbalanced tree-based multilevel discrete wavelet transform was presented.•A new iterative multiple classifiers-based majority voting has been proposed.•A hand-modeled learning model attained over 98% classification accuracy.