基于多尺度粗糙集模型的决策树在高校就业数据分析中的应用

基金项目:

国家自然科学基金项目（61100150）; 广东省高等职业教育教学改革项目（20130201037）; 广东省高职教育信息技术类专业教学指导委员会教学改革项目（XXJS-2013-24）; 广州市属高校科研计划(2012A077)

详细信息

DATA MINING IN EMPLOYMENT BASED ON MULTISCALE ROUGH SET MODEL DECISION TREE

1.
1Department of Electronic and Communication Engineering,Guangdong Industry Technical College,Guangzhou 510300,China;
2.
2Experinental Training Management Center,Guangdong Industry Technical College,Guangzhou 510300,China;
3.
3Department of Computer and Education Software,Guangzhou Univ,Guangzhou 510006,China

摘要: 为解决目前常用于就业数据分析的C4.5算法、基于粗糙集等的决策树生成算法均无法很好地处理决策精度需求不同和噪声适应能力的问题，运用基于多尺度粗糙集模型的决策树算法于于高校就业数据分析，并以某高校2012年就业数据为例进行分析，同时将分析结果与C4.5算法和基于粗糙集的决策树生成算法的分析结果进行比较．结果表明：基于多尺度粗糙集模型的决策树算法生成的决策树树形结构简单、产生的规则简洁、不存在不可分的数据集、运算速度快．
- 多尺度粗糙集模型 /
- 决策树 /
- 就业数据分析 /
- 决策精度 /
- 噪声适应
Abstract: In order to solve different requirements for decision accuracy and noise adaptation, but some decision tree algorithms by C4.5 and base on Rough Set(RS) are used in the analysis of employment data, are unable to solve those problems,this paper present the algorithm to build decision tree based on Multiscale Rough Set Mode into employment data analysis，do a test on employment data of a university in 2012 for training set, and compare results with C4.5 and base on RS. The results show that decision treeconstructed by MRSM is simple, and decision-making rules are Concise, and the data set is end classification, and faster computing speed.
- decision tree /
- multiscale rough set mode /
- employment data /
- decision accuracy /
- noise adaptation