The Design of Desired Collectives with Multi-Agent Simulation

Collectives of Interacting Agents

Agent’s Behavior Based
on the Logic of Minority

Highlights  of The Talk

Slide 5

Logic of Minority: Asymmetric Problems

Reasons for Undesirable Outcomes

Symmetric Problem vs. Asymmetric Problem

 Decomposition to Pair-wise Problems

Desirable Collective:  Stability,  Efficiency, Fairness

Characterization of Learning Models

Slide 12

Coupling of Agents

Coupling Rule between Two

The Performances of Evolutional Learning

What  Agents Acquired with Evolutinary Learning ?

Commonality  of Acquired  Rules

 Coupling with Local Neighbors

Simulation Results

Coupling Agents with Collectives

Simulation Results (θ=0.5)

Efficient Utilization of Limited Resource
with Too Many Contestants

How to Solve Inverse Problem?

Exogenous Design with Subsidy or Tax

Endogenous Design with Give&Take

Slide 26

Conclusion: Achieving Desired Collectives