Augmenting EEG with Generative Adversarial Networks Enhances Brain Decoding Across Classifiers and Sample Sizes.
Chad C. WilliamsDaniel WeinhardtMaria WirzbergerSebastian MusslickPublished in: CogSci (2023)
Keyphrases
- sample size
- small sample
- eeg signals
- statistical tests
- finite sample
- model selection
- brain computer interface
- event related
- brain activity
- random sampling
- confidence intervals
- functional connectivity
- training set
- upper bound
- statistical parametric mapping
- eeg data
- random samples
- naive bayes
- generative model
- training data
- electrical activity
- human brain
- decision trees
- signal processing
- support vector
- brain signals
- event related potentials
- variance reduction
- extracted features
- discriminative classifiers
- statistical power
- supervised classification
- generalization error
- network structure
- class labels
- feature set
- vc dimension
- brain images
- structure learning
- random sample
- statistical significance
- machine learning
- worst case
- supervised learning
- learning algorithm