A study on reward and punishment learning using a data-driven approach

Mental stress is the main well-being problem world-wide today. It is responsible for most of all mental-brain diseases.Mental stress does not need any specific reason to happen. It canbe experienced from a very little incident to a huge issue. Theconsequence of it depends on how people handle it. Depressionand anxiety are one of the results of it, and they are one of themajor challenges in today’s world, and sometimes depression andanxiety patients are taking major steps like suicide. Additionally,most of the time, suicidal patients hide their true feelings andfail to communicate their psychiatric problems to physicians.The specific issues that need to be addressed are finding an easy,reliable and realistic way to diagnose mental stress to keep itfrom becoming a serious and irreversible condition. The primaryprevention of mental stress utilizing machine learning algorithmsbased on reward and punishment processing is important toavoid mental diseases. Several techniques have been used todetect mental stress, and very few papers have tried to detect apatients’ comorbidity condition. However, literature shows thatthere are still chances of further improvement in this field.The traditional methods of detecting Mental stress involve astatistical questionnaire approach with some shortcomings asthey are easy to fake, which is not possible if we use EEG signals.Therefore, in this paper, we proposed a method to evaluate theElectroencephalogram (EEG) signals on thirty-two individualsfor identifying comorbid patients using nine Machine Learningclassifiers based on reward and punishment processing. Theperformance of our method is also shown to be better than someexisting methods.