How deeply does Graph Neural Networks in Action cover heterogeneous graph convolutional network?

Builds a custom MessagePassing layer, relation-specific weights, additive aggregation, and stacked encoder layers; defers mathematical derivation.

Heterogeneous graph convolutional network

As explained in Graph Neural Networks in Action

A graph convolutional network that applies relation-specific transformations or parameters to graphs containing multiple edge types.

Use a heterogeneous graph convolutional network when nodes connect through multiple edge types, such as single, double, triple, or aromatic molecular bonds. It adapts message passing by learning relation-specific parameters instead of treating every edge identically. section 5.4.3

A practical implementation follows this pipeline:

Store one parameter matrix per relation, indexed by each edge's type. section 5.4.3
Run PyG message passing with summation aggregation, passing node features, edge indices, and edge types. section 5.4.3

Molecular graph generation

A variational graph autoencoder can encode molecular graphs with heterogeneous bond types, decode adjacency and node features, and predict a molecular property such as QED.

How Graph Neural Networks in Action covers it

Cover of Graph Neural Networks in Action

liveBook manning.com

Read Graph Neural Networks in Action on liveBook Get the book at manning.com

Related concepts

graph convolutional networkkind of · A graph convolutional network is a graph neural network that computes representations by transforming and aggregating information from graph neighborhoods.
message passinguses · A communication or computation technique in which entities exchange messages to transmit or propagate information, coordinate work, or update state.
heterogeneous graph neural networkkind of · A heterogeneous graph neural network is a graph neural network that learns representations from graphs containing multiple types of nodes, edges, or both.

This explainer was generated by Manning's knowledge graph from the text of Graph Neural Networks in Action. The section links point to the parts of the book it draws on.