Table 1 Classification accuracy and standard deviation (%) of the frustration-level (pBCI) classifier across 12 participants. Overall, the classifier achieved a mean accuracy of 85.2% ± 0.096, demonstrating robust discrimination between low, medium, and high frustration states. As noted in the Results section, this relatively high accuracy may partly reflect the temporal autocorrelation of EEG signals, which can increase apparent separability between conditions. Nevertheless, the results support the reliability of the pBCI classifier as a crucial intermediate component of the proposed framework.

From: Integrating active brain-computer interfaces (aBCIs) with passive BCIs (pBCIs) under different frustration levels