共查询到19条相似文献,搜索用时 46 毫秒
1.
2.
有偏估计与LS估计的比较与选择 总被引:8,自引:0,他引:8
从假设检验的角度研究测量平差Gauss-Markov模型中有偏估计与LS估计的选择问题。首先在均方误差准则下对目前应用最广泛的2种有偏估计--岭估计和主成分估计与LS估计进行了比较研究,得到了岭估计、主成分估计优于LS估计的条件;然后运用统计方法对这些条件的成立进行了假设检验;最后通过数值实验说明,在一定显著性水平下当原假设被接受时,说明没有理由不相信采用有偏估计来代替LS估计的合理性,可认为采用有偏估计将对LS估计做出比较有效的改进,当原假设被拒绝时,说明对采用有偏估计的优越性产生了怀疑,此时建议仍采用LS估计。 相似文献
3.
本文介绍一种求平差参数的有偏估计方法——主成分估计。当法方程系数矩阵呈病态时,主成分估计的均方误差要小于最小二乘估计的均方误差,因此,采用主成分估计是有益的。 相似文献
4.
5.
针对法方程系数阵病态和观测值存在粗差对估值结果造成双重影响的问题,在岭型主相关估计的基础上引入抗差M估计模型,提出了一种抗差有偏估计的新方法——抗差岭型主相关估计。推导出了相应模型的抗差解及误差影响函数,并进行了计算分析。与主成分估计不同的是,该方法按主成分对观测值的影响程度大小对其进行舍取,在尽可能保留有用观测信息的基础上克服了病态性和粗差的影响。 相似文献
6.
针对法方程系数阵病态和观测值存在粗差对估值结果造成双重影响的问题,在岭型主相关估计的基础上引入抗差M估计模型,提出了一种抗差有偏估计的新方法--抗差岭型主相关估计.推导出了相应模型的抗差解及误差影响函数,并进行了计算分析.与主成分估计不同的是,该方法按主成分对观测值的影响程度大小对其进行舍取,在尽可能保留有用观测信息的基础上克服了病态性和粗差的影响. 相似文献
7.
8.
附有条件的参数平差模型的有偏估计 总被引:1,自引:0,他引:1
讨论附有条件的参数平差模型参数的有偏估计问题。提出了约束岭估计和约束主成分估计,并证明了它们的优良性质,最后给出了一个算例,验证了所得结果。 相似文献
9.
抗差主成分估计及应用 总被引:3,自引:0,他引:3
本文基于抗差估计和主成分估计理论,提出一种新的估计-抗差主成分估计,推导了参数的抗差主成分解,并结合算例进行了试算。结果表明:抗差主成分估计不仅可以解除法方程系数降的病态,而且对于抑制观测粗差的影响也有显著功效。 相似文献
10.
11.
Gauss-Markov模型参数的部分根方估计 总被引:3,自引:0,他引:3
在分析根方估计,Stein均压缩估计,Sclove部分压缩估计各自特点的基础上,提出了一种新的有偏估计-部分根方估计,并讨论了它的优良性质。计算结果表明,在设计阵呈病态时,部分根方估计确能改善LS估计。 相似文献
12.
Peter Congdon 《Journal of Geographical Systems》2008,10(3):217-239
To ensure health resources are equitably distributed, composite indices of population morbidity or “health need” are often
used. Measures of the dimensions of population morbidity (e.g. socioeconomic deprivation) relevant to health need are typically
not directly available but indirectly measured through census or other sources. This paper considers measurement of latent
population morbidity constructs using both health outcomes (e.g. hospital admissions, mortality) and observed area social
and demographic indicators (e.g. census data). The constructs are allowed to be spatially correlated between areas, as well
as correlated with one another within areas. The health outcomes may depend both on the latent constructs and on other relevant
covariates (e.g. bed supply), with some covariates possibly measured only at higher (regional) scales. A case study considers
variations in psychiatric admissions in 354 English local authority areas in relation to two latent constructs: area deprivation
and social fragmentation.
相似文献
13.
Latent lifestyle preferences and household location decisions 总被引:3,自引:2,他引:1
Lifestyle, indicating preferences towards a particular way of living, is a key driver of the decision of where to live. We
employ latent class choice models to represent this behavior, where the latent classes are the lifestyles and the choice model
is the choice of residential location. Thus, we simultaneously estimate lifestyle groups and how lifestyle impacts location
decisions. Empirical results indicate three latent lifestyle segments: suburban dwellers, urban dwellers, and transit-riders.
The suggested lifestyle segments have intriguing policy implications. Lifecycle characteristics are used to predict lifestyle
preferences, although there remain significant aspects that cannot be explained by observable variables.
相似文献
Jieping LiEmail: |
14.
Accurate estimates of terrestrial latent heat of evaporation (LE) for different biomes are essential to assess energy, water and carbon cycles. Different satellite- based Priestley-Taylor (PT) algorithms have been developed to estimate LE in different biomes. However, there are still large uncertainties in LE estimates for different PT algorithms. In this study, we evaluated differences in estimating terrestrial water flux in different biomes from three satellite-based PT algorithms using ground-observed data from eight eddy covariance (EC) flux towers of China. The results reveal that large differences in daily LE estimates exist based on EC measurements using three PT algorithms among eight ecosystem types. At the forest (CBS) site, all algorithms demonstrate high performance with low root mean square error (RMSE) (less than 16 W/m2) and high squared correlation coefficient (R2) (more than 0.9). At the village (HHV) site, the ATI-PT algorithm has the lowest RMSE (13.9 W/m2), with bias of 2.7 W/m2 and R2 of 0.66. At the irrigated crop (HHM) site, almost all models algorithms underestimate LE, indicating these algorithms may not capture wet soil evaporation by parameterization of the soil moisture. In contrast, the SM-PT algorithm shows high values of R2 (comparable to those of ATI-PT and VPD-PT) at most other (grass, wetland, desert and Gobi) biomes. There are no obvious differences in seasonal LE estimation using MODIS NDVI and LAI at most sites. However, all meteorological or satellite-based water-related parameters used in the PT algorithm have uncertainties for optimizing water constraints. This analysis highlights the need to improve PT algorithms with regard to water constraints. 相似文献
15.
16.
制图综合方根规律模型的分形扩展 总被引:10,自引:3,他引:10
本文分析了方根规律存在的问题,运用分形理论,研究了有关要素数量指标随比例尺变化的规律,并建立了顾及要素自身形态及分布特征的方根规律分形扩展模型,从理论上,为科学、客观地进行制图综合的数量选取开辟了一条新的途径。本文最后还给出了若干要素选取的具体方法和所提出模型的应用实例 相似文献
17.
18.