Abstract

This project explores techniques for reducing the complexity of an agent-based model (ABM). The analysis involved a model developed from the ethnographic research of Dr. Lee Hoffer in the Larimer area heroin market, which involved drug users, drug sellers, homeless individuals and police. The authors used statistical techniques to create a reduced version of the original model which maintained simulation fidelity while reducing computational complexity. This involved identifying key summary quantities of individual customer behavior as well as overall market activity and replacing some agents with probability distributions and regressions. The model was then extended to allow external market interventions in the form of police busts. Extensions of this research perspective, as well as its strengths and limitations, are discussed.

Highlights

  • We present an example of agent-based model reduction, which is defined as simplifying a detailed agent-based model while preserving the model’s statistical characteristics

  • We evaluated the goodness of fit between the full and reduced models on these statistics to decide whether the fit was adequate

  • Our model reduction procedure explored four versions of the reduced model: one which uses distributions for all five outcomes, one which uses a regression for the probability of obtaining heroin and distributions for everything else, one which uses a regression for the probability of being invited to make direct private deals and distributions for everything else and a final version which uses regressions for the probability of obtaining heroin and the probability of being invited to make direct private deals but distributions for everything else

Read more

Summary

Introduction

We present an example of agent-based model reduction, which is defined as simplifying a detailed agent-based model while preserving the model’s statistical characteristics. The reduced model failed to produce results sufficiently similar to the full model when using this distribution to predict whether customers successfully obtain heroin.

Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call