作者: K. Lugger , D. Flotzinger , A. Schlögl , M. Pregenzer , G. Pfurtscheller
DOI: 10.1007/BF02522476
关键词: Context (language use) 、 Curse of dimensionality 、 Autoregressive model 、 Linear discriminant analysis 、 Dimensionality reduction 、 Pattern recognition 、 Feature extraction 、 Variance (accounting) 、 Mathematics 、 Principal component analysis 、 Artificial intelligence
摘要: The study focuses on the problems of dimensionality reduction by means principal component analysis (PCA) in context single-trial EEG data classification (i.e. discriminating between imagined left- and right-hand movement). components with highest variance, however, do not necessarily carry greatest information to enable a discrimination classes. An set is presented where high variance cannot be used for discrimination. In addition, method based linear discriminant (LDA), introduced that detects which can discrimination, leading sets reduced but similar accuracy.