Research on the measurement and evaluation technique integrated from the viewpoint of ecology to analyze nonlinear information processing in the human behavior

人类的感觉和行为机制是适应复杂环境而形成的高度复杂的非线性信息处理系统。有各种各样的复杂现象。例如..各种行为模式之间的转换，多重意义的竞争性识别，以及生物系统中的混沌动力学等。为了理解人类的这种信息处理过程，将人类系统作为一个复杂的适应系统进行研究变得非常重要。本研究的根本目的是构建人类认知行为的测量体系，整合人类从生物水平到行为和认知水平的各个层次的活动，在本研究中，我们主要围绕以下三个主题展开研究：1.生物非线性动力学分析。自组织信息加工。人体行为的交互测量1)研究了生物内动力学的提取和表征方法。在这里，我们提出了生物动力学的动态实时表征的概念，并将其应用于人体血流动力学的分析。2)研究了多神经智能体系统的自适应学习方法。对于脉冲神经网络，我们提出了时空自组织学习的方法，并将其应用于感觉信息的绑定。3)构建了人机交互系统。本文研究了智能体的行为如何影响人类的心理状态和对智能体的理解，通过这些研究，我们开发了一个基本的框架来处理人类系统的生理或心理因素等非显性的人类信息。

The mechanism of human sense and behavior is a highly complicated nonlinear information processing system that is obtained by adapting to the complex environment. There are various complex phenomena. e.g.. a transition between various behavior patterns, a competitive recognition of multiple meaning, and, chaotic dynamics in biological systems, etc. In order to understand such human information processing, a research of human system, as a complex adaptation system, becomes very important. The fundamental goal of our research is to construct a measurement system of human cognitive behavior and to integrate various levels of human activity from biological level to behavioral and cognitive level, In this research, we mainly focused on the following three subjects as bases of such study :1. Analysis of biological nonlinear dynamics.2. Self-organized information processing.3. Interactive measurement of human behavior.1) The extraction and characterization methods of biological internal dynamics were investigated. Here we developed a concept of dynamic real-time characterization of biological dynamics and applied to the analysis of human blood flow dynamics. 2) The adaptive learning method for multi neural agents system was investigated. For the pulse neural network, we developed the method of spatio-temporal self-organized learning and applied to a binding of sensory information. 3) We constructed the interactive system between human and computer agent. Here we investigated how the behavior of agent influences human mental state and understanding of the agent.Through these studies, we developed a basic framework to handle a non-explicit human information such as physiological or mental factor of human system.