基于模糊模式识别的认知诊断

摘要

目前绝大多数的认知诊断方法都是基于概率统计方法的，需要在大样本的基础上进行复杂的参数估计。本研究采用模糊模式识别的方法对被试进行分类，模拟结果表明，基于格贴近度的被试分类可以得到很高的诊断准确率。实证研究表明，该方法可以对初中二年级学生掌握数学的认知模式进行很好的诊断。该方法原理清楚、计算简便，无需进行复杂的参数估计，在较小样本条件下也可以使用。%At present,most of the cognitive diagnostic tests are based on the statistical methods. They require large sample and complex parameter estimation. In this study,fuzzy pattern recognition was used to classify the subjects. The simulation research showed that assorting based grid closeness has higher diagnostic accuracy. Empirical research demonstrated this method could diagnose the cognitive pattern of eighth graders in mathematical study very well. The principle of this method is quite clear, and the calculation process is simple. It is also applicable when the sample is small.