Читать книгу Artificial Intelligence and Quantum Computing for Advanced Wireless Networks - Savo G. Glisic - Страница 79

In the previous section, we have briefly surveyed a number of GNN designs with the very basic description of their principles. Here, we categorize these options into recurrent graph neural networks (RecGNNs), convolutional graph neural networks (ConvGNNs), graph autoencoders (GAEs), and spatial‐temporal graph neural networks (STGNNs) and provide more details about their operations.

Figures 5.2–5.5, inspired by [38], give examples of various model architectures. In the following, we provide more information for each category, mainly on computation of the state, the learning algorithm, and transition and output function implementations.