Reputation: 1
I recently started learning about GNN algorithms. I am studying GCN algorithm and I want to build an adjacency matrix with time lagged cross correlation instead of Pearson correlation for a data frame with 6 features, can you explain how I should approach it?
It says that cross correlation can only be obtained from two data serials, can't it be applied to the entire data frame with 6 features?
Upvotes: 0
Views: 36