Spatio-spectral & temporal parameter searching using class correlation analysis and particle swarm optimization for a brain computer interface

摘要: Distinct features play a vital role in enabling computer to associate different electroencephalogram (EEG) signals brain states. To ease the workload on feature extractor and enhance separability between states, numerous parameters, such as separable frequency bands, data acquisition channels time point of maximum are chosen explicit each subject. Recent research has shown that using subject specific parameters for extraction invariant characteristics state can significantly improve performance accuracy brain-computer interface (BCI). This paper focuses developing fast autonomous user-specific tuned BCI system Particle Swarm Optimization (PSO) search optimal parameter combination based analysis correlation classes i.e., R-Squared (R2) coefficient rather than assessing overall systems via measure classification accuracy. Experimental results utilizing eight subjects presented which demonstrate effectiveness proposed methods & efficient system.