Invasive Brain Computer Interface (BCI) systems through Electrocorticographic (ECoG) signals require efficient recognition of spatio-temporal patterns from a multi electrodes sensor array. Such signals are excellent candidates for automated pattern recognition through machine learning algorithms. However, the available data is limited due to the operative procedure required for such dataset creation. The importance of different temporal signatures and individual electrodes can be analyzed through feature extraction techniques. But, the variability of the signal due to non-stationarity is ignored while extracting features and which features to use can be challenging to figure out by visual inspection. In this study, we introduce the signal split parameter to account for the variability of the signal, and we use genetic selection, which allows the selection of the optimal combination of features from a pool of 8 different feature sets.