Hypersage

HyperSAGE Layer.

class topomodelx.nn.hypergraph.hypersage.HyperSAGE(in_channels, hidden_channels, n_layers=2, alpha=-1, **kwargs)

Neural network implementation of HyperSAGE [1] for hypergraph classification.

Parameters:

in_channelsint

Dimension of the input features.

hidden_channelsint

Dimension of the hidden features.

n_layersint, default = 2

Amount of message passing layers.

alphaint, default = -1

Max number of nodes in a neighborhood to consider. If -1 it considers all the nodes.

**kwargsoptional

Additional arguments for the inner layers.

References

[1]

Arya, Gupta, Rudinac and Worring. HyperSAGE: Generalizing inductive representation learning on hypergraphs (2020). https://arxiv.org/abs/2010.04558

forward(x_0, incidence_1)

Forward computation through layers, then linear layer, then global max pooling.

Parameters:

x_0torch.Tensor, shape = (n_nodes, features_nodes)

Edge features.

incidence_1torch.Tensor, shape = (n_nodes, n_edges)

Boundary matrix of rank 1.

Returns:

torch.Tensor, shape = (1)

Label assigned to whole complex.