graph perceptron

Data

Graph Perceptron

Let $σ : R \to R$ . The graph perceptron is defined as

\begin{aligned} x \in R^{n} & \to u = \sum_{k = 0}^{K - 1} h_{k} S^{k} x graph conv \to_{u} \\ \to_{u} y \cdot σ (u) pointwise nonlinearity \to_{y} \end{aligned}

ie a graph convolution followed by a pointwise nonlinear function

y_{i} = σ (u_{i})

Examples

File
GNNs inherit stability from their layers
multi-layer graph perceptron
2025-02-05 graphs lecture 5
2025-03-10 graphs lecture 13