Upstairs, we can determine whether there are edges between the nodes of the brain functional network according to sparsity. For example, the sparsity is set to 0.2, that is to say, the number of edges existing in the current brain functional network accounts for 20% of the number of edges in the complete network (calculated by the complete graph). The specific method is as follows: arrange the correlation matrices in descending order of correlation coefficient values, and take the first 20% correlation coefficient (the position of these correlation matrices is set to 1), and the position of other correlation matrices is set to 0. Sparsity is a variable, generally from 0.05 to 0.5.