改进的ID3算法在远程教学系统中的应用  

Application of improved ID3 algorithm to distance education system

作　　者：王建伟[1] 王鑫[1] 许宪东[1] 李慧君[1] 黑龙[1] 

机构地区：[1]黑龙江工程学院计算机科学与技术学院,黑龙江哈尔滨150050

出　　处：《黑龙江工程学院学报》2014年第1期67-70,共4页Journal of Heilongjiang Institute of Technology

基　　金：黑龙江省自然科学基金项目(F201224)

摘　　要：当前,远程教学系统缺少智能性,不能提供个性化教学,引入ID3算法后可以根据学习者的特征对其分类,从而实现对不同学习者的针对性教学。然而传统的决策树ID3算法存在多值倾向的问题,选择分裂属性不符合客观事实。运用一种基于灰色关联分析的修正因子属性选择方法予以改进,对取值较多但灰色关联度低的属性,在计算其信息增益时通过灰色关联度的正弦值作为修正因子,克服传统ID3算法的不足。将改进的ID3算法引入到远程教学系统中,可以更好地对学习者进行分类以实现智能化导学。Currently,the distance education system is lack of intelligence and cannot provide personalized teaching which can be classified according to the feature of different learners after introducing ID3 algorithm in order to realize the targeted teaching to different learners. There are multi-valued problems in the traditional decision tree algorithm ID3, and split attribute selecting does not conform to the objective facts. A feature selection method of correction factor is used based on grey relational analysis. The grey correlation sine value is selected as a correction factor when calculating the information gain for the properties of low value but more grey correlation degree to overcome the deficiency of the traditional ID3 algorithm. The introduction of improved ID3 algorithm can classify learners better to achieve intelligent tutoring.

关 键 词：ID3算法 决策树 灰色关联度 修正因子 远程教学系统 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]