This paper is devoted to the results of a study on the classification of human electroencephalogram
signals to determine the intention to make a movement. Signals were subjected to artifact removal and filtering, and then biomarkers responsible for self-initiated movements were extracted from them. Signal classification was performed using machine learning methods. The quality of the different models was compared, the best of which was demonstrated by a support vector machine. The classification results can be used to develop control algorithms based on neurofeedback.