| 引用本文: | 邢涛,廖冉.多分类支持向量方法的本科招生生源质量模型[J].哈尔滨工业大学学报,2010,42(11):1828.DOI:10.11918/j.issn.0367-6234.2010.11.032 |
| XING Tao,LIAO Ran.Multi-category support vector machine algorithm for quality analysis of student enrollment of universities[J].Journal of Harbin Institute of Technology,2010,42(11):1828.DOI:10.11918/j.issn.0367-6234.2010.11.032 |
|
| 摘要: |
| 为了建立一套科学、有效、完善的招生管理流程和评价体系,对多分类支持向量机算法进行改进,利用各地区学生在大学期间的学习成绩与学生入学前的信息特点,探讨本科生源质量预测体系,建立了本科生源质量分析模型.结果表明,改进的多分类支持向量机能够较好地应用于高校本科招生生源质量分析研究,建立的招生生源质量模型可以对本科生源质量进行良好的预测,为高校制定招生计划、提高生源质量提供有效的参考.通过对惩罚矩阵进行合理的调整,模型学习的结果和可靠性都有显著提高。 |
| 关键词: 数据挖掘 多分类支持向量机 统计学习理论 招生 质量分析 |
| DOI：10.11918/j.issn.0367-6234.2010.11.032 |
| 分类号:G647 |
| 基金项目: |
|
| Multi-category support vector machine algorithm for quality analysis of student enrollment of universities |
|
XING Tao1, LIAO Ran2
|
|
1.School of Economics and Management,Beijing University of Aeronautics and Astronautics,Beijing 100191,China;2.School of Mathematics and Systems Science,Beijing University of Aeronautics and Astronautics,Beijing 100191,China)
|
| Abstract: |
| In order to explore and establish a scientific effective and comprehensive recruitment management process and valuation system,and based on several information before students entering college,such as regions and location,the score in college entrance examination,the rank in certain regions,a new model to deal with enrollment strategy and related matters has been set up in this paper.By a real case,it shows that the model is suitable to analyze the quality of fresh students,and can provide proper guidance for planning college enrollment stralegy.The model reliability and result accuracy can be significantly inproved by reasonable adjustment of penalty matrix. |
| Key words: data mining multi-category support vector machine statistical learning theory students’sources enrollment analysis of quality |
