Evaluation of mother wavelets on steady-state visually-evoked potentials for triple-command brain-computer interfaces

Wavelet transform (WT) is an important tool to analyze the time-frequency structure of a signal. The WT relies on prototype signal that called mother wavelet. However, there no single universal wavelet fits all signals. Thus, selection function might be challenging represent achieve optimum performance. There are some studies determine optimal for other biomedical signals; however, exists evaluation steady-state visually-evoked potentials (SSVEP) signals becomes very popular among manipulated brain-computer interfaces (BCIs) recently. This study aims explore, if any, suits best SSVEP classification purposes in BCIs. In this study, three common wavelet-based features (variance, energy, and entropy) extracted from five distinct EEG frequency bands (delta, theta, alpha, beta, gamma) were classified different user commands using six fundamental classifier algorithms. was repeated commonly-used functions (haar, daubechies, symlet, coiflet, biorthogonal, reverse biorthogonal). discrimination obtained with accuracy 100% average 75.85%. Besides, ensemble learner gives highest accuracies half trials. Haar had performance representing wavelets adopted study. Concomitantly, variance, entropy should used together since none these superior alone.