| 基于密度分布的半监督回归算法研究 |
| |
| 引用本文： | 张倩,李明,王雪松. 基于密度分布的半监督回归算法研究[J]. 工矿自动化, 2012, 38(3): 29-30 |
| |
| 作者姓名： | 张倩 李明 王雪松 |
| |
| 作者单位： | 中国矿业大学信电学院,江苏徐州,221116 |
| |
| 基金项目： | 国家自然科学基金项目(61072094,60804022);江苏省2011年度普通高校研究生科研创新计划项目(CXLX11-0318) |
| |
| 摘 要： | 提出了推导密度函数的基本假设,对密度函数进行了推导,通过密度函数实现了密度区域的划分;对同一密度范围内的未标签值标记的估计给出了具体的处理方法;最后介绍了基于密度分布的半监督回归算法的具体实现步骤。该算法实现了对未标签点的标记,能够减小对未标签点标签值的估计误差,提高估计的准确度。
|
| 关 键 词： | 机器学习 半监督学习 回归算法 密度分布 |
Research of Semi-supervised Regression Algorithm Based on Density Distribution |
| |
| ZHANG Qian , LI Ming , WANG Xue-song. Research of Semi-supervised Regression Algorithm Based on Density Distribution[J]. Industry and Automation, 2012, 38(3): 29-30 |
| |
| Authors: | ZHANG Qian LI Ming WANG Xue-song |
| |
| Affiliation: | ( School of Information and Electrical Engineering of CUMT.,Xuzhou 221116,China) |
| |
| Abstract: | The paper proposed a basic hypothesis for deducing density function and derived density function.It achieved density regional division through the density function,gave a specific estimating approach for computing tag value of untagged label within the same density range,and described specific implementation steps of semi-supervised regression algorithm based on density distribution.The algorithm achieves mark of untagged points,and can reduce estimating error of the tag value and improve estimation accuracy. |
| |
| Keywords: | machine learning semi-supervised learning regression algorithm density distribution |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
