notesum.ai
Published at December 6Feature-Based Network Construction: From Sampling to What-if Analysis
physics.soc-ph
cs.SI
Released Date: December 6, 2024
Authors: Christian Franssen1, Joost Berkhout2, Bernd Heidergott1
Aff.: 1Dept. Operations Analytics, VU Amsterdam, The Netherlands; 2Dept. of Mathematics, VU Amsterdam, The Netherlands

| Feature function | Symbol | Description | Definition |
| In-strength | The sum of link weights reaching each node . | (1) | |
| Out-strength | The sum of link weights leaving each node . | (2) | |
| Reciprocity | The extent to which nodes reciprocate incoming | Appendix A.1 | |
| links. | |||
| Triangular | The extent to which nodes form closed triplets. | Appendix A.2 | |
| closures | |||
| Modularity | The extent to which nodes are grouped in | Appendix A.3 | |
| communities or clusters. | |||
| Assortativity | Tendency for nodes to link with others that are | Appendix A.4 | |
| similar in node covariates or a function over the | |||
| network itself (e.g., node strengths). | |||
| Stationary | The long-run fraction of time the random walker | Appendix A.5 | |
| spends at each node . | |||
| Kemeny | The expected time to reach a -randomly drawn | Appendix A.6 | |
| node from any arbitrary node. | |||
| Effective graph | The sum of electrical resistances in a resistor | Appendix A.7 | |
| resistance | network with edge weights as conductances. |