Content deleted Content added
m v2.04 - Repaired 1 link to disambiguation page - Bonds / Fix errors for CW project (Square brackets without correct end - Reference before punctuation - Reference tags without correct match - Unbalanced quotes in ref name or illegal character.) |
NickDiCicco (talk | contribs) m Corrected wrong capitalisation in a subscript |
||
Line 96:
:<math>\mathbf{X}' = (\mathbf{X} \odot \text{sigmoid}(\mathbf{y}))_{\mathbf{i}}</math>
:<math>\mathbf{A}' = \mathbf{A}_{\mathbf{i}, \mathbf{
Where <math>\mathbf{X}</math> is the matrix of node features, <math>\mathbf{A}</math> is the graph adjacency matrix, <math>\odot</math> denotes element-wise [[matrix multiplication]], and <math>\mathbf{p}</math> is a learnable [[Projection (mathematics)|projection]] vector. The projection vector <math>\mathbf{p}</math> computes a scalar projection value for each graph node. The nodes with the top-k highest projection scores are retained in the new adjacency matrix <math>\mathbf{A}'</math>. The <math>\text{sigmoid}(\cdot)</math> operation makes the projection vector <math>\mathbf{p}</math> trainable by [[backpropagation]], which otherwise would produce discrete outputs.<ref name=gao2019 />
| |||