1. Background
GCNs differ from CNNs
- Each node with a different number of adjacent nodes has a filter of different sizes and weights
- the weights in the filters are the same for all neighboring nodes in the receptive field.
weights are not learned
Related Work
Graph attention networks (GATs) employed the attention mechanism to obtain different and
trainable weights for adjacent nodes by measuring the correlation between their feature vectors
and that of the central node.
extra computation in terms of pairs of feature vectors
评论0