Application of Neural Networks in Studying of Curricula Correlation
- 物理技术－已发表论文 
用神经网络的自组织学习算法 ,对某班学生的语文、数学、英语、物理、化学这 5个科目的成绩进行无导师学习 ,观察网络的分类结果 .发现语文和数学两种课程的关联程度最低 ,说明它们是两门不同类型的学科 .其结论对于中小学的课程设计和安排具有一定的参考价值 .A Self Organizing Mapping (SOM) network was designed and trained by the scores of the students of a class in a senior high school in the examinations of five courses (i.e., Mathematics, Chinese, English, Physics and Chemistry). From the classification results of the network, It is found that the distance between the winning neuron of Mathematics and that of Chinese to be the remotest. This implies that Mathematics and Chinese are independent and have weak correlativity. The results is referencable in curricula designing and planning.