Identifying the Neurocognitive Difference Between Two Groups Using Supervised Learning

• Activity Number: 210 - SLDS CSpeed 3
• Type: Contributed
• Date/Time: Tuesday, August 10, 2021 : 1:30 PM to 3:20 PM
• Sponsor: Section on Statistical Learning and Data Science
• Author(s): Ramchandra Rimal*
• Companies: Middle Tennessee State University
• Keywords: fMRI data; brain imaging ; supervised learning; statistical network modeling

Abstract

Brain Imaging Analysis is among one of the most exciting fields in neuroscience. We aim to develop the classification model to improve the predictive performance and analyze the compromise between the accuracy and the interpretability of the model. We work with the chess masters dataset to identify the neurocognitive differences between chess masters and chess novice groups. For this, we analyze the resting-state fMRI data of the two groups and create a network of the connections between brain regions. We then develop a classification framework implementing statistical network modeling techniques to distinguish the connectivity patterns between the groups. In particular, we develop two classification frameworks, one for interpretability and another optimized for precision. We then analyze the trade-off between precision and the interpretability of the model. We report the classification accuracy measures to justify the performance of our framework.