Lecture

• during the practicals
• 30. 11. – 4. 12. (1. 12. or 2. 12.)

we can pass the exam (but it won't be noted if we don't)

• find relationships among the data
• visualize the links and relationships

• some nodes (= hubs) have extremely many connections to other nodes
  • most nodes have just a few connections to other nodes
  • resilience to random disturbance
  • vulnerability to attacks
  • → new application areas (malware protection, medicine, business)
  • different than random graphs (normal distribution of the number of edges)

- random failures: 80 % of randomly selected nodes can fail without framenting the cluster
- coordinated attacks: the elimination of 5–15 % of all the hubs can cause the entire system to collapse