ID3算法在程序设计类课程成绩分析中的应用研究  

Application of the decision tree analysis technique in scores of programming course

作　　者：刘敏娜[1,2] 

机构地区：[1]咸阳师范学院信息工程学院,陕西咸阳712000 [2]咸阳师范学院图形图像处理研究所,陕西咸阳712000

出　　处：《电子设计工程》2016年第9期42-44,47,共4页Electronic Design Engineering

基　　金：陕西省教育厅专项基金资助项目(15JK1803);陕西省科学技术研究发展计划项目(2013JM8037);咸阳师范学院专项科研基金项目(14XSYK036)

摘　　要：基于分析学生成绩主要影响因素的目的,采用了ID3算法完成决策树的生成。算法包括数据采集、数据处理、绘制决策树、决策树剪枝、去噪和模型准确性评估等阶段。模型选取软件工程专业2012级1班的学生的《JAVA程序设计》课程期末考试成绩和学生的基本情况信息等数据作为训练数据。经过反复训练,模型能够根据输入学生的基本信息分析出学生的学习成绩优,良,差的概率。该决策树准确率为90%,能够满足用户需求,决策树模型符合要求。Based on the analysis of the main influence factors of studentsˊ performance, the ID3 algorithm is used to accomplish the decision tree. Algorithm consists of data acquisition, data processing, drawing decision tree, decision tree pruning, to noise and the accuracy of the model evaluation. Model selection of software engineering 1 class 2012 students of the ＆quot;JAVA program design＆quot; course final exam results and studentsˊ basic information and other data as training data. After repeated training, the model can be based on the basic information of the students to analyze the studentsˊ learning performance, good, poor probability. The accuracy rate of the decision tree is 90%, which can satisfy the needs of users, and the decision tree model meets the requirements.

关 键 词：ID3算法 MATLAB 决策树 数据挖掘 

分 类 号：TN9[电子电信—信息与通信工程]