Comprehensive Analysis on Information Transmission among Agents

在这项研究中，我的目标是在综合数据的收集、计算系统、数学技术和实地研究之间发展良好的平衡方法。具体来说，我试图开发方法，以基于大量数据和基于智能体的建模的智能体行为因果关系为基础，从整体的角度捕捉智能体之间的通信状态。我收集了外汇市场全面的高分辨率数据（2007年5月至2011年1月，45个月），并构建并行计算系统进行存储和计算。此外，我开发了通过报价和交易的发生来捕获代理通信结构的方法，并基于这些方法实现了计算机软件。在整个调查期间，我们遭受了几次与全球经济危机有关的极端事件，如最近的全球金融危机（2008年7月至2009年2月）和欧元冲击（2010年4月至5月）。经证实，所提议的方法能够捕捉到这些危机，并发现与大量市场参与者有关的影响。与大量人口相关的同步活动可能会推动宏观经济结构的变化。

In this study, I aimed to develop well-balanced methodologies among collection of comprehensive data, a computational system, mathematical technique, and a field study. Specifically, I attempted to develop methods to capture states of communication among agents from a holistic point of view based on a large amount of data and agent-based modeling regarding causality of agent behavior.I collected comprehensive high-resolution data of the foreign exchange market (May 2007 to January 2011; 45 months) and constructed the parallel computing system in order to store and calculate them. Further, I developed methods to capture structure of agent communication through occurrence of quotations and transactions and implemented computer software based on the methodologies. Throughout the period of this investigation, we suffered from several extreme events relating to economic crises across the globe such as the latest global financial crisis (July 2008 to February 2009), and the Euro shock (April to May 2010). It was confirmed that the proposed methods can capture these crises and detect influence relating to a large number of market participants. It is concluded that synchronous activities relating to a large number of population may drive the change of macroeconomic structure.