EdgeRegression

class graphstorm.model.EdgeRegression(h_dim, target_etype, out_dim=1, dropout=0, norm=None, use_bias=True)

Bases: GSEdgeDecoder

Decoder for edge regression tasks.

Can be used for edge regression tasks or edge feature reconstruction.

New in version 0.4.0: The EdgeRegression.

Parameters

h_dim: int: The input dimension size.
target_etype: tuple of str: The target etype for prediction in the format of (src_ntype, etype, dst_ntype).
out_dim: int: The output dimension size. Default: 1.
dropout: float: Dropout rate. Default: 0.
norm: str, optional: Normalization methods. Not used, but reserved for complex edge regression. implementation. Default: None.
use_bias: bool: Whether the edge decoder uses a bias parameter. Default: True.

forward(g, h, e_h=None)

MLP-based edge decoder forward computation.

g: DGLGraph: The graph of target edges.
h: dict of Tensor: The input node embeddings in the format of {ntype: emb}.
e_h: dict of Tensor: The input edge embeddings in the format of {(src_ntype, etype, dst_ntype): emb}. Not used, but reserved for future support of edge embeddings. Default: None.

predict(g, h, e_h=None)

MLP-based edge regression prediction computation.

g: DGLGraph: The graph of target edges.
h: dict of Tensor: The input node embeddings in the format of {ntype: emb}.
e_h: dict of Tensor: The input edge embeddings in the format of {(src_ntype, etype, dst_ntype): emb}. Not used, but reserved for future support of edge embeddings. Default: None.

predict_proba(g, h, e_h=None)

MLP-based edge regression prediction computation. It returns the same results as the predict() function.

g: DGLGraph: The graph of target edges.
h: dict of Tensor: The input node embeddings in the format of {ntype: emb}.
e_h: dict of Tensor: The input edge embeddings in the format of {(src_ntype, etype, dst_ntype): emb}. Not used, but reserved for future support of edge embeddings. Default: None.

property in_dims: Return the input dimension size, which is given in class initialization.