On the Selection of Adaptive Algorithms in ABM: A Computational-Equivalence Approach

Abstract

Abstract. Agent-based Methodology (ABM) is becoming indispensable for the interdisciplinary study of social and economic complex adaptive systems. The essence of ABM lies in the notion of autonomous agents whose behavior may evolve endogenously and can generate and mimic the corresponding complex system dynamics that the ABM is studying. Over the past decade, many Computational Intelligence (CI) methods have been applied to the design of autonomous agents, in particular, their adaptive schemes. This design issue is non-trivial since the chosen adaptive schemes usually have a profound impact on the generated system dynamics. Robert Lucas, one of the most influential modern economic theorists, has suggested using laboratories with human agents, also known as Experimental Economics, to help solve the selection issue. While this is a promising approach, laboratories used in the current experimental economics are not computationally equipped to meet the demands of the selection task. This paper attempts to materialize Lucas' suggestion by establishing a laboratory where human subjects are equipped with the computational power that satisfies the computational equivalence condition.

Computationally intelligent agents in economics and finance

“…This is basically what was done in [46], which estimated the interaction mechanism among young people in relation to smoking behavior by using the result of [2]. 15 Its empirical results strongly support the presence of positive peer effects in smoking behavior among young people.…”

“…There are a series of agent-based financial econometric models which do not explicitly refer to a network or a graph [43,33,53]. 15 Ref. [2] can be read as one of the earliest available econometric results of agent-based models.…”

Assessing the economic ripple effect of flood disasters in light of the recovery process: Insights from an agent‐based model

“…ABM is becoming significant in the study of economic complex adaptive systems (Chen & Tai, 2006). The ripple effect is a macroemergent result of the economic impact of flood risk.…”

Economía artificial: una valoración crítica

“…En nuestra opinión, resulta más útil desagregar esta disciplina en subdisciplinas estructuralmente más próximas. Aparte de las áreas ya mencionadas, desde el punto de vista metodológico se consideran dentro de la economía computacional la economía computacional basada en agentes o ACE (Tesfatsion y Judd, 2006;Tesfatsion, 2003), el desarrollo de herramientas de programación específicas para la economía (Kendrick y Amman, 1999) o las aplicaciones de la inteligencia artificial en economía y gestión, entre otras muchas (Chen et al, 2006); desde el punto de vista de la aplicación a áreas económicas concretas, se incluyen, por ejemplo, las finanzas computacionales y el modelado computacional de sistemas macroeconómicos dinámicos (Amman et al, 1996;Schmedders y Judd, 2014).…”