From: Predicting miRNA-disease associations via layer attention graph convolutional network model
Input: |
miRNA–miRNA similarities matrix, disease–disease similarities matrix, adjacent matrix A; |
Construct the input graph G = (V,E) |
Output: |
Initialize embedded dimension, learning rate, training epoch, dropout rates; |
Initialize embeddings; |
Iterate according to the layerwise propagation rule of GCN; |
Introduce an attention mechanism; |
Combine other embeddings and obtain final embeddings of miRNAs and diseases; |
Obtain the Loss function |