“…In the application of evolutionary algorithms to improve the performance of BMI-based systems, the three evolutionary algorithms of GA, PSO and DE are most commonly used. Based on the conducted studies, the application of the PSO algorithm in BMI can be categorizes in three groups of filter design [164], feature selection, frequency and optimal electrodes [59,167,168,171,175,176,190], and classification [9,69,177,178,[180][181][182][183][184]191]. [164] uses five generalized versions of PSO (constant weighted inertia-PSO, linear decay inertia-PSO, constriction factor inertia-PSO, nonlinear inertia-PSO, and dynamic inertia-PSO) in order to design the optimal coefficients of the adaptive noise canceller (ANC) [192] for filtering of EEG signals.…”