Tikhonov regularized spectrally weighted common spatial patterns

Common spatial patterns(CSP) is one of the most popular feature extraction methods used in Brain-computer interfaces(BCI). Different variants of Common spatial patterns exist. One of the best among them is Spectrally weighted common spatial patterns(SPEC-CSP). In this paper we introduce Tikhonov regularized spectrally weighted common spatial patterns(TRSPEC-CSP) and Weighted Tikhonov regularized spectrally weighted common spatial patterns(WTRSPEC-CSP). The proposed methods are not so sensitive to noise and not prone to over fitting. We used dataset 2a of BCI competition IV to evaluate our method. The result shows that proposed methods outperforms SPEC-CSP and the best competitor of the BCI competition IV.