基于神经网络和可拓学的高中生解决矛盾问题研究

Neural Networks and Extenics Based Contradiction Problem Solving Study Of High School Students

运用自组织神经网络SOM,将学生划分为3个不同学习成绩的组别。考察他们在解决矛盾问题时所采用的算子,并运用方差分析和多重比较方法进行差异显著性检验。结果表明,高中生能够采用多种可拓算子来解决矛盾问题,学习成绩和可拓算子的使用间存在复杂的关系,某些可拓算子的使用在三组学生间并无显著性差异,某些可拓算子的使用在两组学生之间没有显著性差异。

Students were clustered into 3 groups according their achievements with SOM networks. The operators of contradiction problem solving were reviewed. Significant tests were conducted with ANOVA and multiple comparing. It was showed that high school students could solve contradiction problems with multi - method. There were complex relationship between achievement and using extension operators. There was no significant difference for some extension operators among three groups, or between two groups.