| 半参数模型估计的补偿最小二乘法与正则核方法 |
| |
| 引用本文： | 李环波,刘立龙,金惠国.半参数模型估计的补偿最小二乘法与正则核方法[J].测绘科学,2011,36(3):158-159. |
| |
| 作者姓名： | 李环波 刘立龙 金惠国 |
| |
| 作者单位： | 桂林理工大学,广西桂林,541004 |
| |
| 基金项目： | 国家自然科学基金项目(40574002); 广西青年科学基金项目(0728097); 广西研究生教育创新计划项目(2008105960816M20); 广西区应用基础研究专项(桂科基0991023) |
| |
| 摘 要： | 本文根据测量数据处理中利用半参数对系统误差去除的效果,提出了半参数补偿最小二乘法和正则核方法两种方法进行比较.两种去除系统误差的方法都存在着局限性,应该根据系统误差的具体情况,选择不同的方法;最后通过模拟算例去除系统误差,并与实测数据进行拟合验证两种方法的可行性.
|
| 关 键 词： | 半参数估计 正则化 补偿 系统误差 |
Penalized least squares method and the kernel method of Semi-parametric model estimation |
| |
| LI Huan-bo,LIU Li-long,JIN Hui-guo.Penalized least squares method and the kernel method of Semi-parametric model estimation[J].Science of Surveying and Mapping,2011,36(3):158-159. |
| |
| Authors: | LI Huan-bo LIU Li-long JIN Hui-guo |
| |
| Institution: | LI Huan-bo,LIU Li-long,JINHui-guo(Guilin Institute of Technology,Guangxi Guilin 541004,China) |
| |
| Abstract: | Based on the effect of systematic error removal using semi-parameters in measuring data processing,the paper compared between semi-parametric penalized least squares method and the kernel method.Because both methods exists limits,different methods should be selected depending on practical cases.The paper finally gave simulation examples to remove systematic errors,and fit with actual measurement to validate the feasibility of two methods. |
| |
| Keywords: | semi-parametric estimation regularization compensation system error |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
