SAGEConv

GraphSAGE convolution layer (Hamilton et al., 2017). $\mathbf{x}_i' = \mathbf{W}_1 \mathbf{x}_i + \mathbf{W}_2 \cdot \text{mean}_{j \in \mathcal{N}(i)} \mathbf{x}_j$

Record fields

Name	Description
`Bias`
`WeightNeighbor`
`WeightSelf`

Instance members

Name	Description
`this.forward (x, edgeIndex)`