论文信息 - A method for determining a classifier for determining the state based on electroencephalography data

A method for determining a classifier for determining the state based on electroencephalography data

A method (300) for determining a state classifier described. The method (300) comprises determining (304) a plurality of training data sets. Furthermore, the method (300) comprises determining (305, 306), based on the plurality of training data sets weighting values for individual source features, in which a weighting value for a source feature a relevance of this source characteristic for the status indicates classifier, and wherein the weighting values are determined by reducing or increasing an optimization function. The optimization function in this case comprises a first elimination term, the rewards it if all source characteristics of a particular channel (405) a plurality of different channels (405) of an electroencephalograph (105) for the state classifier are irrelevant. Furthermore, the optimization function comprises a second elimination term, the rewards, when a single origin feature for the state classifier is irrelevant. It can be selected based on the weighting values a relevant subset of the plurality of source characteristics for the state classifier.

[1] Roberto Muñoz-Guerrero,et al. Analysis and classification of electroencephalographic signals (EEG) to identify arm movements , 2013, CCE.

[2] Ramaswamy Palaniappan,et al. Neural network classification of autoregressive features from electroencephalogram signals for brain–computer interface design , 2004, Journal of neural engineering.

[3] Chin-Teng Lin,et al. Classification of driver's cognitive responses from EEG analysis , 2005, 2005 IEEE International Symposium on Circuits and Systems.