RATIONAL ANALYSIS OF LEARNING AND COGNITION

学习和认知的理性分析

• 批准号: 3075865
• 负责人: JOHN Robert ANDERSON
• 金额: $ 9.93万
• 依托单位: CARNEGIE-MELLON UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 1989
• 资助国家: 美国
• 起止时间: 1989-06-01 至 1994-05-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/3075865
• 关键词: cognition; computer simulation; learning; problem solving

The principle methodology of this research has been to develop computer simulations of the representations and mechanisms of human cognition. My research has taken a new direction which is characterized as being focused on how the mechanisms of the human mind are adapted to structure of the environment. I am currently pursuing a thesis called "The principle of rationality" which proposed that the mechanisms of the mind are optimized to the environment. A book is being completed (Anderson, in press) that develops this thesis successfully for relatively simple areas of human cognition such as basic memory and simple concept formation. The purpose of this award is to allow sufficient time to acquire the training required to effectively pursue this analysis for the more complex areas of problem solving and language. The first three years will be spent acquiring a sophisticated background in Bayesian decision theory. The last three years (overlapping in one year) will be spent retraining myself as to the current state of linguistics. In parallel with this, a rational theory of problem solving will be developed. This theory will combine an analysis of the probability of causal extrapolations involved in problem solving with the cost of various paths of problem solving. It will attempt to explain both the logic and control of problem solving as an optimization that responds to the combined consideration of probability and cost.

这项研究的主要方法是发展 计算机模拟人类的表现和机制 认知. 我的研究有了新的方向， 其特点是专注于人类的机制如何 我们的思想适应了环境的结构。 我目前 追求一个名为“合理性原则”的论文， 他提出，大脑的机制是最优化的， 环境 一本即将完成的书（安德森，出版中）， 发展这篇论文成功地为相对简单的领域， 人类认知，如基本记忆和简单概念的形成。 这一裁决的目的是为了有足够的时间获得 有效地进行这种分析所需的培训， 更复杂的问题解决和语言领域。 第一 三年的时间将用来获得一个复杂的背景， 贝叶斯决策理论 过去三年（重叠在一个 一年）将用于重新培训自己，以了解目前的状态， 语言学 与此同时，解决问题的理性理论将 发展。 该理论将联合收割机与 问题解决中的因果推断概率 与解决问题的各种途径的成本。 它将试图 把解决问题的逻辑和控制力解释为一种 优化，响应于以下综合考虑 概率和成本。