Network Science for Fun and Profit
summary
Describes using R + igraph and other tools (NetworkX, Gephi, Neo4j, GraphChi/GraphLab) to fit real networks to models—random, small‑world, scale‑free—and to run SI/SIR diffusion simulations (small‑world saturates ~9 steps, scale‑free ~28). Shows community detection with walktrap community detection and a practical rebuild: hubs are reaggregated so device clusters map to individuals, then a deployed trust score predicts fraud successfully after reaggregation; reports graph stats like density ≈0.16 and hub degrees (regular ≈4 vs scale‑free ≈40).
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:11Talk goals and framework
- 2:59What is network science
- 10:03Graph types and representations
- 16:24Graph generative models
- 21:45Analyzing graph properties
- 23:29Case study: Florentine marriages
- 30:31Understanding centrality
- 45:58Network models in real world
- 53:07Community detection
- 1:00:27Network diffusion and epidemics
- 1:07:56Next steps and resources
- 1:15:12Q&A
related talks
24mHands-On Network Science
46mGraph Algorithms: Predict Real-World Behavior
02h 47mNetwork Analysis Made Simple
43mVisualising networks
28mA new wave of Network Science - from fantasy to applications
01h 26mBuilding an Interactive Network Graph to Understand Communities
01h 27mNetwork Science, Game of Thrones and US Airports
01h 32mConnected: A Social Network Analysis Tutorial with NetworkX