Evaluating Stress through Machine Learning based on Brain Activity Data

University essay from KTH/Skolan för elektro- och systemteknik (EES); KTH/Skolan för elektro- och systemteknik (EES)

Author: Christian Agnér; Anneli Blomqvist; [2017]

Keywords: ;

Abstract: More people are experiencing stressrelatedsymptoms, which is not only causing worsenhealth, but also causing economical drawbacks for thesociety, businesses and individuals. The aim of thisproject is to create a tool that evaluates stress frombrain activity data and can help to avoid develop thesymptoms.An EMOTIV Epoch EEG headset is used to recorddata. The stress level is evaluated from the brainactivity data by the parameters, feeling of pleasure(valence) and the mental workload. k-NN machinelearning is utilized to create a valence classificationalgorithm and the theta power density spectrum is usedto determine the workload. Eye movement disturbancesin the recordings are also addressed.Tests with Stroop word color games as stress stimuliare conducted and the project concludes that it ispossible to determine the stress level correctly, onaverage, 17.56% and when allowing one level difference,48.71% .

  AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)