Complex Systems and Agent Based Modelling

Complex systems:

• Many interacting parts (complex, interwoven)
  
• The whole is greater than the sum of its parts
  
• Not reducible to its parts
• Not predictable
  
• Nonlinear
  
• Self organising
  
• Path dependent
  
• Emergent
  
• Chaotic
  
• Positive and Negative Feedback
  
• High Dimensionality
  
• Power Law

Economy, social systems, markets, healthcare

Complexity:

Social Events, Economic Depressions, Crises, and Organisational Change

Why Agent Based Models?

Current approaches:

• Linear Thinking
  
• Top-down
  
• Reductionist
  

are not good enough
We need Complexity Thinking

• Nonlinear
  
• Bottom-up
  
• Multi disciplinary

Agent Based Modelling / Simulation:

• Computational approach
  
• Autonomous agents
  
• Bottom-up modelling
  
• Simulate and observe
  
• Look for emergence and unexpected
  
• Educational
  
• Visualise and create virtual labs for discussion
  
• Generate insights, understand actors