A service provided by the WU Library and the WU IT-Services

A generic simulation environment for heterogeneous agents. With applications in marketing and technological choice.

Meyer, David (2003) A generic simulation environment for heterogeneous agents. With applications in marketing and technological choice. Doctoral thesis, WU Vienna University of Economics and Business.

[img]
Preview
PDF
Download (823Kb) | Preview

Abstract

This monograph contributes to the methodology of Agent-based Computational Economics. First, we introduce a generic simulation framework suitable for agent-based simulations featuring the support of heterogeneous agents, hierarchical scheduling, and flexible specification of design parameters. For the latter, we use an XML-based format enabling the design of flexible models, with the possibility of varying both agent population and parameterization. Further, the tool allows the definition of communication channels to single agents, or groups thereof, and handles the information exchange. Both agents and communication channels can be added and removed at runtime. To handle the heterogeneity arising from both the agents' implementations and the underlying platforms, we introduce an XML-based wrapper technique for mapping the functionality of agents living in an interpreter-based environment to a standardized JAVA interface. Second, we present a collection of artificial economic actors to be used with this framework. Their interplay is demonstrated in two fields of management science: marketing and technological choice. In the field of marketing, the question of choosing the optimal segmentation techniques for market segmentation is investigated, comparing the performance of firm agents with diverse segmentation strategies in a highly customizable artificial consumer market. In the second application, we study the influence of technological efficiency and organizational inertia on the emergence of competition when firms decide myopically. We observe the competitive reaction of a former monopolist to the advent of a new competitor to assess when new, "disruptive" technologies cause the failure of incumbent firms and investigate simple defensive strategies. (author's abstract)

Item Type: Thesis (Doctoral)
Keywords: agents / simulation / simulation framework / JAVA / disruptive technologies / market segmentation / simulation control / experimental design / R / wrapper / simulation manager / XML
Divisions: Departments > Finance, Accounting and Statistics > Statistics and Mathematics
Depositing User: Repository Administrator
Date Deposited: 06 Oct 2003 12:25
Last Modified: 27 Apr 2015 15:52
URI: http://epub.wu.ac.at/id/eprint/1904

Actions

View Item