作者: RN Khushaba , A Al-Jumaily , A Al-Ani
DOI:
关键词:
摘要: One of the most important tasks in any pattern recognition system is to find an informative, yet small, subset features with enhanced discriminatory power. In this paper, a new neuro-fuzzy discriminant analysis based feature projection technique presented on two stages hybrid Neural Networks, optimized Differential Evolution (DE), and proposed Fuzzy Linear Discriminant Analysis (FLDA) technique. Although dimensionality reduction via FLDA can present set well clustered reduced space, but like version existing DA’s it assumes that original data linearly separable, which not case many real world problems. order overcome problem, first stage maps initially extracted nonlinear manner into domain, larger dimensionality, are separable. acts then these separable further reduce dimensionality. The combination, referred as NFDA, validated prosthetic device control problem Electroencephalogram (EEG) datasets collected from 5 subjects achieving maximum testing accuracy 85.7% for three classes EEG imaginations movements. Keywords—Feature Projection, Analysis, Prostheses Control.