Applying Decision Tree Algorithm on Gamified Data of Continuous Performance Test for Detection of Hyperactivity / Attention Deficit Disorder.

Abstract

The three main goals of this study are the evaluation of Decision Tree algorithm for detecting ADHD disorder, reduction of the number of required features, and providing acceptable prediction despite the lack of several features. The subjects in this study were 100 people betwee the ages of 8 and 64 years, 69 of whom were under 18 and 31 were over 18. Of these, 43 have ADHD, 35 are healthy and 22 are suspected. The data was collected by the Brain and Cognition Clinic in Tehran using the AIV-2 game. The results indicate that applying machine learning methods are still effective even with involvement of the least number of features in algorithm implementation process.