torch_geometric.explain.metric.characterization_score

characterization_score(pos_fidelity: Tensor, neg_fidelity: Tensor, pos_weight: float = 0.5, neg_weight: float = 0.5) → Tensor[source]

返回如“GraphFramEx: Towards Systematic Evaluation of Explainability Methods for Graph Neural Networks”论文中所述的组件特征评分。

\[\textrm{charact} = \frac{w_{+} + w_{-}}{\frac{w_{+}}{\textrm{fid}_{+}} + \frac{w_{-}}{1 - \textrm{fid}_{-}}}\]
Parameters:

  • pos_fidelity (torch.Tensor) – 正保真度 \(\textrm{fid}_{+}\).

  • neg_fidelity (torch.Tensor) – 负保真度 \(\textrm{fid}_{-}\).

  • pos_weight (float, optional) – \(\textrm{fid}_{+}\) 的权重 \(w_{+}\)。（默认值：0.5）

  • neg_weight (float, optional) – \(\textrm{fid}_{-}\) 的权重 \(w_{-}\)。（默认值：0.5）

Return type:

Tensor