A method for evaluating the degree of human's preference based on EEG analysis

We always employ a subjective evaluation for recognizing user's preference. However, it depends on the day or the person. It is difficult to evaluate the human's preference objectively, but it is needed for developing in the factory. By the way, sense of touch is important factor for human to decide what they like. In this paper, we propose a method for extracting an information on the sense of touch based on electroencephalogram (EEG) analysis. We analyze the EEG signals while we touch objects. We often use the frequency analysis for data analysis of the EEG. The frequency analysis is used in much work of EEG. There seems to be frequency bands reflected in the preference significantly. Therefore, we apply particle swarm optimization (PSO) for selection of the significant component. For separating, support vector machine (SVM) is used. Experimental result shows that the error rates of the proposed method indicate lower value. Therefore, proposed method has the capability of separating human's preference.