A Functional Connectivity-Based EEG Analysis for the Decision-Makings in Ultimatum Game

In the ultimatum game (UG), individuals usually reject extremely unfair offers and accept fair ones. A mid-value offer is unfair, but not to the point of immediate rejection. When faced with this, most individuals hesitate to either accept or reject. As a result, there is rich decision-making information in the brain during mid-value offers, and recognizing whether individuals will accept or reject in this situation is challenging. In recent years, EEG channel connectivity indices have been widely used to study brain networks. Among them, mutual information (MI), phase-locked value (PLV), and phase-lag index (PLI) are commonly employed to construct functional connectivity matrices (FCMs). Since these functional connectivity indices are undirected, the FCMs constructed from them are symmetric matrices with a large amount of redundant information. To address this, this paper utilizes a variety of connectivity indices and constructs a fused functional connectivity matrix by pairing two of them. The results show that the approach using PSI + MI performs optimally in the classification task of recognizing whether an individual accepts or rejects and reliably extracts a network structure that reveals inter-channel connection strength variations in decision-makings,