Stationary Subspace Analysis as a pre-processed Adaptive Common Spatial Pattern Data Method

Abstract

Brain activity is described by brain waves. Retrieving brain wave data can use a special recording device. The results of data taken on the record tool sometimes have bad noise. Bad noise is caused by interference received by the subject during data retrieval. The Common Spatial Pattern (CSP) method is a method for separating features from brain waves. The Common Spatial Pattern method has experienced very rapid development. Adaptive Common Spatial Pattern (ACSP) is a development method of CSP. CSP development becomes ACSP due to CSP's inability to handle data with many subjects. Previous ACSP research carried out the application of ACSP using raw data. The application of the raw data is to determine the accuracy of ACSP. In this study, we will use the stationary subspace analysis (SSA) method as a data preprocessing method. The results of this study indicate that the SSA method can increase the results of 1% accuracy.