Link Prediction algorithm

[himil48]

Hi

I would like to perform link prediction between Patient and Condition nodes. I followed code provided in a tutorial posted on medium that implements the following model.

class GNNEncoder(torch.nn.Module):
    def __init__(self, hidden_channels, out_channels):
        super().__init__()
        self.conv1 = SAGEConv((-1, -1), hidden_channels)
        self.conv2 = SAGEConv((-1, -1), out_channels)

    def forward(self, x, edge_index):
        x = self.conv1(x, edge_index).relu()
        x = self.conv2(x, edge_index)
        return x


class EdgeDecoder(torch.nn.Module):
    def __init__(self, hidden_channels):
        super().__init__()
        self.lin1 = Linear(2 * hidden_channels, hidden_channels)
        self.lin2 = Linear(hidden_channels, 1)

    def forward(self, z_dict, edge_label_index):
        row, col = edge_label_index
        z = torch.cat([z_dict['user'][row], z_dict['movie'][col]], dim=-1)

        z = self.lin1(z).relu()
        z = self.lin2(z)
        return z.view(-1)

class Model(torch.nn.Module):
    def __init__(self, hidden_channels):
        super().__init__()
        self.encoder = GNNEncoder(hidden_channels, hidden_channels)
        self.encoder = to_hetero(self.encoder, data.metadata(), aggr='sum')
        self.decoder = EdgeDecoder(hidden_channels)

    def forward(self, x_dict, edge_index_dict, edge_label_index):
        z_dict = self.encoder(x_dict, edge_index_dict)
        return self.decoder(z_dict, edge_label_index) 

Are there perhaps any other models that could be recommended for executing this task.
Thank you,

[wsad1]

Check out the following examples

1. Link pred on movie lens(heterogenous graph): link
2. Link pred examples on homogenous graphs, which can give you ideas on new ecoders/decoders: link1, link2.

[himil48]

Thank you, I will give that a try.