This is a preliminary study about which of two classifiers: Support vector machine (SVM) or linear discriminant analysis (LDA), and which frequency band: δ (0.1-4Hz), μ (8-12Hz) and β (6-31Hz), provide higher accuracy using brain-computer interface (BCI) for detecting two different cognitive states: Pedaling (a motor complex imagery task) and relaxation. Results show that after using independent components analysis, in δ band for 3 out of 5 subjects achieved over 90% of accuracy and the other two over 60% of accuracy.
Rodriguez-Ugarte, M., Angulo-Sherman, I. N., Ianez, E., Ortiz, M., & Azorin, J. M. (2018). Preliminary study of pedaling motor imagery classification based on EEG signals. In 2017 International Symposium on Wearable Robotics and Rehabilitation, WeRob 2017 (pp. 1–2). Institute of Electrical and Electronics Engineers Inc. https://doi.org/10.1109/WEROB.2017.8383851