A unified account of learning, memory, and categorization

学习、记忆和分类的统一描述

• 批准号: 355882-2008
• 负责人: Jamieson, Randall
• 金额: $ 1.23万
• 依托单位: University of Manitoba
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2008
• 资助国家: 加拿大
• 起止时间: 2008-01-01 至 2009-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=402752
• 关键词: unified; account; learning; memory; categorization

What is a memory? How does a person's wealth of prior experience come to influence the decisions they make? Can an understanding of human memory lead to the design of intelligent machines? My research program is aimed at answering these questions. To achieve this goal my students and I develop computer models that describe the mathematics of learning and categorization. We test the models by matching the behaviour of the models against that of people. By representing memory as a store of specific experiences, and retrieving those memories in parallel, we have explained how people become sensitive to regularities in the world, how they will choose to sort items into categories, and how they will conceive of categories flexibly under different learning circumstances.

什么是记忆？一个人之前丰富的经验是如何影响他们所做的决定的？对人类记忆的理解能导致智能机器的设计吗？我的研究计划就是为了回答这些问题。为了实现这一目标，我和我的学生开发了描述学习和分类数学的计算机模型。我们通过将模型的行为与人的行为相匹配来测试模型。通过将记忆描述为特定经验的存储，并并行地检索这些记忆，我们解释了人们如何对世界的规律变得敏感，他们如何选择将物品分类，以及他们如何在不同的学习环境下灵活地构思类别。