INDUCTION OF CATEGORY DISTRIBUTIONS - A FRAMEWORK FOR CLASSIFICATION LEARNING
INDUCTION OF CATEGORY DISTRIBUTIONS - A FRAMEWORK FOR CLASSIFICATION LEARNING
复制标题
DOI：
10.1037/0278-7393.10.2.234
复制
发表时间：
1984-01-01
影响因子：
2.6
通讯作者：
HOLYOAK, KJ
中科院分区：
文献类型：
--
作者：
FRIED, LS;HOLYOAK, KJ
提出了一个分类学习的框架，假设学习者使用所提供的实例（无论是标记或未标记）来推断密度函数的类别样本在一个特征空间，随后的分类决策采用相对似然决策规则的基础上，这些推断的密度函数。一个具体的模型，基于这个一般框架，类别密度模型，提出了考虑到诱导的正态分布的类别，无论有或没有错误校正或提供标记的实例。该模型被实现为计算机模拟。对257名大学生进行的5个实验结果表明，被试不仅在不纠错的情况下，而且在不知道类别数量甚至不知道有类别需要学习的情况下，都能学习类别分布。研究结果指示一个更一般的学习模型，集成分布表示的基础上参数描述和存储的实例。(38参考）（PsycINFO数据库记录（c）2016阿帕，保留所有权利）
Presents a framework for classification learning that assumes that learners use presented instances (whether labeled or unlabeled) to infer the density functions of category exemplars over a feature space and that subsequent classification decisions employ a relative likelihood decision rule based on these inferred density functions. A specific model based on this general framework, the category density model, is proposed to account for the induction of normally distributed categories either with or without error correction or provision of labeled instances. The model was implemented as a computer simulation. Results of 5 experiments with 257 undergraduates indicated that Ss could learn category distributions not only without error correction, but also without knowledge of the number of categories or even that there were categories to be learned. Findings dictate a more general learning model that integrates distributional representations based on both parametric descriptions and stored instances.(38 ref)(PsycINFO Database Record (c) 2016 APA, all rights reserved)