Estimating Human Attentional Focus based on EEG signals using Machine Learning

抄録

Attentional focus plays a crucial role in various cognitive processes, impacting performance in work, study, and many activities. Numerous studies explored the relationship between EEG signals and attentional focus to gain a deeper understanding of this cognitive phenomenon. However, estimating attentional focus still poses challenges in the scientific field, due to the non-linearity and individual differences of EEG signals. In recent years, the employment of machine learning for classification of EEG signals increased in many fields due to the robustness of the models. Building on this, we propose to estimate attentional focus by employing machine learning based on the EEG signals and the performance of participants during mental activities. We conducted an experiment, wherein participants wore an EEG sensor and underwent a prolonged cognitive test to induce attentional focus. We labeled the acquired data as optimal and unfavorable attentional focus based on the performance of participants on the test and used it to train a support vector machine classifier. Although future improvements are required, our methodology achieved an overall accuracy of 82%.