Towards Classification and Prediction of Stress Patterns using Multiple Physiological Signals

Abstract

The stress is increasing in our society in the last years, due the large and tiring routines besides few time to rest. Keeping this in mind, this paper intends to determine patterns in stress’ events using physiological signs, because these signals are a reliable source to identify stress states. The literature shows that the use of physiological signs as a source for stress patterns identification is a promising investigation subject and there are few studies evaluating the effect of combining several different signals. The objective of this article is to investigate the possible integration of data obtained from electrocardiographic (ECG), electrodermal activity (EDA) and electromyography (EMG) to detect stress patterns using wearable sensors to acquisition of biofeedback and propose algorithms to set some patterns. It was developed a dataset to made the pre-processing in all of data to evaluate the plausibility and develop an adequate database for the application of machine learning techniques establishing as a reference the obtained annotated data.