This paper presents a methodological approach to designing agent-based models (ABMs), emphasizing the step-by-step design process that supports model development. Using real estate investment dynamics in urban environments as a case study, the paper outlines a framework for developing an ABM. The design guidelines provided are informed by literature and experiential insights gained through the development of our own model, offering a practical roadmap for ABM creation. This framework focuses on key system design elements: mapping the system (understanding and empathizing with it), conducting feasibility analysis (determining what can be done), and performing impact analysis (how it can be achieved). The framework follows an iterative process, allowing for continuous model evolution and refinement. This framework employs a bottom-up approach to designing an agent-based model, starting by defining the micro-level behaviour of agents based on literature, expert insights, and foundational equations. The resulting effects are then observed at the macro-level, allowing for analysis of system-wide phenomena. The core contribution of this paper lies in its focus on the design methodology, offering a versatile approach that can be adapted by those new to the field of agent-based modelling or with limited experience. It provides guidance to help beginners improve their understanding of the design process and successfully implement agent-based models in their own systems. Future studies will focus on validating the design guidelines with expert input and aligning the model’s results with real-world data.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Designing Agent-Based Models: A Methodological Approach from Practical Insights and Lessons Learned from Case Study

  • Vighneshkumar Rana,
  • Vishal Singh

摘要

This paper presents a methodological approach to designing agent-based models (ABMs), emphasizing the step-by-step design process that supports model development. Using real estate investment dynamics in urban environments as a case study, the paper outlines a framework for developing an ABM. The design guidelines provided are informed by literature and experiential insights gained through the development of our own model, offering a practical roadmap for ABM creation. This framework focuses on key system design elements: mapping the system (understanding and empathizing with it), conducting feasibility analysis (determining what can be done), and performing impact analysis (how it can be achieved). The framework follows an iterative process, allowing for continuous model evolution and refinement. This framework employs a bottom-up approach to designing an agent-based model, starting by defining the micro-level behaviour of agents based on literature, expert insights, and foundational equations. The resulting effects are then observed at the macro-level, allowing for analysis of system-wide phenomena. The core contribution of this paper lies in its focus on the design methodology, offering a versatile approach that can be adapted by those new to the field of agent-based modelling or with limited experience. It provides guidance to help beginners improve their understanding of the design process and successfully implement agent-based models in their own systems. Future studies will focus on validating the design guidelines with expert input and aligning the model’s results with real-world data.