pygod.nn.encoder

Personalized Neural Network Encoders

class pygod.nn.encoder.GNA(in_channels, hidden_channels, num_layers, out_channels, dropout=0.0, act=<function relu>)

Graph Node Attention Network (GNA). See [YZY+21] for more details.

Parameters:

• in_channels (int) – Input dimension of node features.

• hidden_channels (int) – Hidden dimension of the model.

• num_layers (int) – Number of layers in the model.

• out_channels (int) – Output dimension of the model.

• dropout (float, optional) – Dropout rate. Default: 0..

• act (callable activation function or None, optional) – Activation function if not None. Default: torch.nn.functional.relu.

forward(s, edge_index)

Forward computation.

Parameters:

• s (torch.Tensor) – Input node embeddings.

• edge_index (torch.Tensor) – Edge index.

Returns:

s – Updated node embeddings.

Return type:

torch.Tensor