The Study of Emotional Brain to Detect Emotions Using Brain EEG Signals and Improving Accuracy of Emotion Detection System Using Feature Selection Techniques

Now a days Emotion detection using brain EEG signal is becoming interest area of many researchers because of it's tremendous application in healthcare and BCI field. Database acquisition, pre-processing, feature extraction and classification are the main stages in this process. In this research study first existing database of brain EEG signal are studied. Most of the researchers used DEAP database for emotion classification. DEAP database is especially made for music recommendation system. Because of the non-linear and non- stationary nature and poor spatial resolution of Brain EEG signals, researchers faced challenges in each phase of emotion detection process. It is found that the classification accuracy is very low. It becomes necessary to study emotional brain and according to that select electrodes for emotion detection to improve classification accuracy. In this research study self-created dataset is used. Two way approach is used for feature selection to improve accuracy. In the first approach least correlated features are omitted from feature set. and in the second approach RFE recursive feature elimination technique is used for feature ranking. The features ranked high are considered in feature set. It is found that classification accuracy is improved using these techniques.