Main Article Content
BCI is one of the interesting areas of research from engineering, psychology, rehabilitation, neuroscience ,computer science and physiology. Providing communications capabilities to highly disabled people like totally paralyzed or ‘locked in’ by neuromuscular neurological disorders like amyotrophic lateral sclerosis, brain stem, stroke, spinal cord injury ; is the important goal of BCI. It is well known that a traditional technique has lowest performance as compare recent technologies in case of detection of stress or any psychological problem. Here, in this paper explained brain interface technique for detection /analysis of stress. The investigation aims to find out best combination of algorithm and classifier which is resulted in highest accuracy for recorded input based on alpha and beta waves. The average sensitivity, average specificity and F-scores are the parameters considered for the analysis purpose . The studied the k-NN, SVM, NB, CT and NN classifier in combination with KDE, RER, Hjorth, ELC and BFCC feature extraction algorithm. It was observed that the combination of k-NN classifier along with ELC features extraction algorithm gave highest percentage accuracy for happiness up to 89.03%. Similarly, we checked the accuracy for happy, sad, clamp and angry emotions.