Session to Session Transfer Learning Method Using Independent Component Analysis with Regularized Common Spatial Patterns for EEG-MI Signals


Training the user in Brain-Computer Interface (BCI) systems based on brain signals that recorded usingElectroencephalography Motor Imagery (EEG-MI) signal is a time-consuming process and causes tiredness to thetrained subject, so transfer learning (subject to subject or session to session) is very useful methods of training that willdecrease the number of recorded training trials for the target subject. To record the brain signals, channels orelectrodes are used. Increasing channels could increase the classification accuracy but this solution costs a lot of moneyand there are no guarantees of high classification accuracy. This paper introduces a transfer learning method usingonly two channels and a few training trials for both feature extraction and classifier training. Our results show that theproposed method Independent Component Analysis with Regularized Common Spatial Pattern (ICA-RCSP) willproduce about 70% accuracy for the session to session transfer learning using few training trails. When the proposedmethod used for transfer subject to subject the accuracy was lower than that for session to session but it still better thanother methods.