Targeted predictability measures for specific network types

Investigate whether more targeted approaches exist to characterize predictability for specific classes of networks, including temporal networks, directed networks, and hypergraphs, and ascertain their suitability for these distinct structural settings.

Background

The paper notes that different domains and network structures (e.g., temporal, directed, and higher-order/hypergraph networks) often display distinct topological properties and evolutionary mechanisms, suggesting that a one-size-fits-all predictability metric may be inadequate.

This motivates the need to explore specialized, structure-aware measures that can more accurately characterize predictability in these heterogeneous network types.