EEG data classification method based on graph representation learning model

The invention belongs to the field of graph embedding, and discloses an EEG data classification method based on a graph representation learning model, which comprises the following steps: step 100, constructing a geodesic distance matrix and an adjacent matrix generated based on spectral coherence of power spectral density characteristics; step 200, calculating a self-adaptive adjacency matrix according to the geodesic distance matrix and an adjacency matrix generated based on the spectral coherence of the power spectral density characteristic, and acquiring the power spectral density of the EEG data as an input node characteristic; and step 300, performing multi-channel attention convolution on the brain function network to obtain final model parameters, and realizing EGG classification. According to the method, node features are coarsened by adopting a global attention pooling mechanism, and embedding of the whole graph is realized. And according to a whole graph classification task, using a constructed los