吉林大学学报(理学版)

• 计算机科学 • 上一篇    下一篇

基于改进蜂群算法的项目反应理论混合模型参数估计

温长吉1,2, 王生生2, 赵昕1, 李健1, 周翠娟1   

  1. 1. 吉林农业大学 信息技术学院, 长春 130118; 2. 吉林大学 符号计算与知识工程教育部重点实验室, 长春 130012
  • 收稿日期:2013-12-09 出版日期:2014-07-26 发布日期:2014-09-26
  • 通讯作者: 王生生 E-mail:wss@jlu.edu.cn

Modified ABC Based Parameter Estimation ofthe Mixture IRT Model

WEN Changji1,2, WANG Shengsheng2, ZHAO Xin1, LI Jian1, ZHOU Cuijuan1   

  1. 1. College of Information and Technology, Jilin Agricultural University, Changchun 130118, China;2. Key Laboratory of Symbolic Computing and Knowledge Engineering of Ministry of Education,Jilin University, Changchun 130012, China
  • Received:2013-12-09 Online:2014-07-26 Published:2014-09-26
  • Contact: WANG Shengsheng E-mail:wss@jlu.edu.cn

摘要:

基于混合项目反应理论模型, 实现了对客观题组和主观测试组成项目结构的模型构建, 并给出一种改进人工蜂群算法实现该混合模型的参数估计. 在模拟数据上通过与Winbug的对比结果验证了利用改进人工蜂群算法估计混合模型参数的有效性.

关键词: 项目反应理论, 混合模型, 参数估计, 人工蜂群算法

Abstract:

A mixture IRT model was proposed which is composed of the testlet response model and the partial credit model for representing the subject and object tests. Meanwhile a modified artificial bee colony algorithm was proposed to estimate the parameters in the mixture model. The comparative experiment results based on the simulate data show that the novel method in this work is feasible and effective.

Key words:  item response theory, mixture model, parameter estimation, artificial bee colony algorithm

中图分类号: 

  • TP301.6