EEG Pattern Classification of Picking and Coordination Using Anonymous Random Walks

Abstract

Tacit coordination games are games where players are trying to select the same solution without any communication between them. Various theories have attempted to predict behavior in tacit coordination games. Until now, research combining tacit coordination games with electrophysiological measures was mainly based on spectral analysis. In contrast, EEG coherence enables the examination of functional and morphological connections between brain regions. Hence, we aimed to differentiate between different cognitive conditions using coherence patterns. Specifically, we have designed a method that predicts the class label of coherence graph patterns extracted out of multi-channel EEG epochs taken from three conditions: a no-task condition and two cognitive tasks, picking and coordination. The classification process was based on a coherence graph extracted out of the EEG record. To assign each graph into its appropriate label, we have constructed a hierarchical classifier. First, we have distinguished between the resting-state condition and the other two cognitive tasks by using a bag of node degrees. Next, to distinguish between the two cognitive tasks, we have implemented an anonymous random walk. Our classification model achieved a total accuracy value of 96.55%.

Offline EEG hyper-scanning using anonymous walk embeddings in tacit coordination games

“…Each experimental session consisted of twelve tacit coordination games, each with a different set of four words (see Appendix A in [ 9 ]). For example, game board #1 displays the set {"Water", "Beer", "Wine", "Whisky"} appearing in Hebrew.…”

“…To be able to produce an optimal classification process, we would like to represent the graph in a way that will preserve its spatial properties, which have been proven to be critical for classifying brain processes associated with the spatial structure of the EEG data [ 9 , 19 , 20 ]. The random walk embedding method assumes that collecting all the single random walks into a probability density function, out of all possible walks, will describe the spatial structure of the graph regardless of the specific node labels where the walks were performed.…”

Predicting Tacit Coordination Success Using Electroencephalogram Trajectories: The Impact of Task Difficulty

“…In one of our recent studies [3], we designed a method that predicts the class label of coherence graph patterns associated with different cognitive states extracted out of multi-channel EEG epochs. Furthermore, in our subsequent study [8], our proposed algorithm found an automated solution such that by analyzing spatial data using random walks [9][10][11], we were able to dramatically improve the classification of the task-related classes associated with precision and recall by ~89% and ~90%, respectively.…”

“…The EEG data underwent several preprocessing steps, which were successfully implemented in previous studies (e.g., [3,8]). These steps included applying band-pass filtering (BPF) in the range of [1,32] Hz to capture the delta, theta, alpha, and beta frequency bands.…”

Enhancing EEG-based attachment style prediction: unveiling the impact of feature domains

“…Our approach aligns with current trends in neuroscience and psychology, where ensemble machine-learning methods have shown effectiveness in interpreting complex neural patterns ( Rahman et al, 2022 ; Li et al, 2023 ). Our study is focused on examining a broad range of EEG features, totaling 45 in number, which encompass elements from the time domain ( Al-Fahoum and Al-Fraihat, 2014 ; Zuckerman et al, 2022 , 2023b ), frequency-based analyses ( Mizrahi et al, 2022a , b , 2023a ), and complexity measures ( Sheehan et al, 2018 ; Ramadoss et al, 2022 ; Mizrahi et al, 2023b ). The aim is to utilize these features to predict whether an individual has a secure or insecure attachment style and to assess the specific contribution of each feature to this prediction.…”