Web proceedings papers


Filip Georgiev and Nevena Ackovska


The use of electroencephalography (EEG) for emotion recognition has gathered a lot of attention in the scientific community. Accurate recognition and understanding of human emotions can improve cooperation between humans and machines. If such a system were established, it would increase productivity in many segments, with special emphasis on medicine, education, the military, etc. Various techniques in machine learning are constantly evolving and are promis-ing new discoveries in this area of research and can be used to process brain signal data. This study attempts to reaffirm some of the existing discoveries in this field of work while working with publicly available data. Several models were being created by different algorithms, where an XGBoost classifying model achieved the best result with 99.94% accuracy, outperforming the models created in the similar studies working with the same dataset.


emotions, EEG, machine learning, XGBoost, LSTM