Stress Condition Detector Based on the Physiological Response using K-Means Cluster Analysis

Everyone is required to be able to solve problems in living daily life that can cause stress. From various studies on stress, most state that stress contributes negatively to health, and increases the risk of disease and decreased endurance. This makes early stress monitoring very necessary, by identifying stress conditions can help in preventive steps to manage stress symptoms to improve and avoid the impact of prolonged stress that affects health such as sleep disorders, changes in appetite, and even one’s social life. From changes in the body’s physiological response can be known stress conditions through heart rate, body temperature and galvanic skin response (GSR). Based on this, this study made a portable tool to detect stress conditions based on physiological changes in the body then a decision was made for someone in normal, mild stress, or severe conditions using K-Means Cluster Analysis. The results of stress condition detection are displayed on the android UI via Bluetooth connection. Test results from 45 datasets obtained centroid points on the sensor for each stress level. Sensor reading values that are close to the centroid value will be grouped in the centroid cluster.