Classical BCI classifiers (CSP-LDA, MDRM) give better calibrated uncertainty than deep learning, which tends to be overconfident, but deep learning remains more accurate.
Robust mo- tor imagery tasks classification approach using bayesian neural network,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Uncertainty Quantification for Motor Imagery BCI -- Machine Learning vs. Deep Learning
Classical BCI classifiers (CSP-LDA, MDRM) give better calibrated uncertainty than deep learning, which tends to be overconfident, but deep learning remains more accurate.