随机变量多重Weibull统计模型及参数最优估计

利用Weibull分布函数兼容性好的特点,采用混合Weibull分布模型来描述复杂随机变量分布,并按照最小二乘法原则,对统计模型参数进行最优化估计,建立复杂随机变量分布的统计模型。对某汽车发动机工作转矩统计模型的建立示例表明,多重Weibull分布适合用来表达一些复杂随机变量的统计特性,参数估计的最优化模型适用,参数优化估计过程能可靠收敛。     

应用模糊聚类评价布氏田鼠为害等级

由于各地环境条件和布氏田鼠密度与种群分布格局的差异,其为害程度很难用统一的数量化指标加以评价,造成了准确测报该鼠为害水平的困难。本文试采用模糊聚类的方法,结合该鼠数量、分布格局以及多项环境因子的定性、定量指标加以综合分析,依其聚类结果可进行该鼠为害等级的全面评价,根据作者收集的20组数据的微机处理,表明这种方法可进行大区域不同环境条件下的为害等级评价,具有良好的应用前景。     

Comparison of methods to assess dog owners' therapeutic compliance

Objective To compare different methods for assessing the compliance of veterinary clients administering medication to their dogs.
Procedure Thirty-one owners whose dogs were prescribed amoxycillin-clavulanate, twice and thrice daily, for a duration of five to seven days were recruited from three Sydney veterinary hospitals. Compliance was assessed by electronic monitoring devices, return medication counts, client self-reports and veterinarians' estimation of likely compliance.
Results Electronic monitoring showed owners administered on average 84% (range 7 to 104%) of prescribed medication to their dogs. Both return medication counts and client self-reports tended to overestimate therapeutic compliance when compared with electronic monitoring. When questioned, the majority of owners (71%) claimed perfect compliance with the prescribed regimen. No correlation was found between veterinarians' estimates of owner compliance and that assessed electronically.
Conclusion Electronic monitoring provided valuable information on dose timing and variation, but proved costly and difficult to employ in veterinary practice. Simplicity and low cost of return medication counts makes this method attractive for use in veterinary compliance studies. Client self-reports reliably detected some noncompliers and permitted identification of individual problems or errors. For practical purposes a combination of return medication counts and client self-reports may have merit in future veterinary compliance studies.     

Estimation of the carcass composition from a cross-section of the rib-loin from crossbred Japanese Black × Limousin F2 cattle by computer image analysis

The carcass composition of crossbred Japanese Black × Limousin F2 cattle was examined in order to find an accurate carcass composition equation. The test animals included 17 steers and 17 heifers. The 28 image measurements from the area encircling the vertical line to the thoracic vertebra and the line from the thoracic vertebra between the sixth and seventh rib‐bones were measured by computer image analysis. The relationships between the 29 parameters that added the carcass left side weight of the animal and the carcass composition were suggested. The carcass composition included muscle weight, muscle ratio, fat weight and fat ratio. The carcass composition from steers was estimated by an equation composed of these three or four parameters (R² = 90.80%, 79.30%, 90.75% and 73.70%, respectively). The selected parameters were measured without cutting the thoracic vertebra. The carcass composition from heifers was estimated by an equation composed of two to four parameters (R² = 96.15%, 90.98%, 93.60% and 88.22%, respectively). The parameters for the estimation of the muscle and fat weight, and muscle and fat ratio are very similar. Furthermore, the equations using the parameters could estimate the carcass composition from the Japanese Black × Limousin cattle resource population.     

Factors Associated with Grassland Bird Species Richness: The Relative Roles of Grassland Area, Landscape Structure, and Prey

The factors responsible for widespread declines of grassland birds in the United States are not well understood. This study, conducted in the short-grass prairie of eastern Wyoming, was designed to investigate the relationship between variation in habitat amount, landscape heterogeneity, prey resources, and spatial variation in grassland bird species richness. We estimated bird richness over a 5-year period (1994–1998) from 29 Breeding Bird Survey locations. Estimated bird richness was modeled as a function of landscape structure surrounding survey routes using satellite-based imagery (1996) and grasshopper density and richness, a potentially important prey of grassland birds. Model specification progressed from simple to complex explanations for spatial variation in bird richness. An information-theoretic approach was used to rank and select candidate models. Our best model included measurements of habitat amount, habitat arrangement, landscape matrix, and prey diversity. Grassland bird richness was positively associated with grassland habitat; was negatively associated with habitat dispersion; positively associated with edge habitats; negatively associated with landscape matrix attributes that may restrict movement of grassland bird; and positively related to grasshopper richness. Collectively, 62% of the spatial variation in grassland bird richness was accounted for by the model (adj-R² = 0.514). These results suggest that the distribution of grassland bird species is influenced by a complex mixture of factors that include habitat area affects, landscape pattern and composition, and the availability of prey.     

青藏高原东部雨季降水量分布模型的建立

采用高分辨率的3"数字高程模型资料和青藏高原东部102个常规气象观测站5～9月份的降水资料,根据降水随高度分布将站点分为3类,采用多元逐步回归的方法,建立了青藏高原40年(1961-2000年)雨季逐年降水量和经度、纬度、海拔高度、坡度、坡向、开放度等地理、地形因子之间的关系模型,估算了青藏高原地区降水量的空间分布.结果表明:建立的关于青藏高原降水量与诸因子之间方程的相关性显著,相对误差在20%,平均相对误差在4.4%之内,估算模型的相关系数均通过置信度为0.95 的检验;并且地理地形因子对40年平均和逐年降水量分布特征的影响较为稳定,其归一化的模拟系数基本控制在±1之间,这对定量分析和评价站点稀少、地形复杂的高原降水有重要意义.     

Density estimates of rural dog populations and an assessment of marking methods during a rabies vaccination campaign in the Philippines

We estimated the population density of dogs by distance sampling and assessed the potential utility of two marking methods for capture-mark-recapture applications following a mass canine rabies-vaccination campaign in Sorsogon Province, the Republic of the Philippines. Thirty villages selected to assess vaccine coverage and for dog surveys were visited 1 to 11 days after the vaccinating team. Measurements of the distance of dogs or groups of dogs from transect lines were obtained in 1088 instances (N = 1278 dogs; mean group SIZE = 1.2). Various functions modelling the probability of detection were fitted to a truncated distribution of distances of dogs from transect lines. A hazard rate model provided the best fit and an overall estimate of dog-population density of 468/km² (95% confidence interval, 359 to 611). At vaccination, most dogs were marked with either a paint stick or a black plastic collar. Overall, 34.8% of 2167 and 28.5% of 2115 dogs could be accurately identified as wearing a collar or showing a paint mark; 49.1% of the dogs had either mark. Increasing time interval between vaccination-team visit and dog survey and increasing distance from transect line were inversely associated with the probability of observing a paint mark. Probability of observing a collar was positively associated with increasing estimated density of the dog population in a given village and with animals not associated with a house. The data indicate that distance sampling is a relatively simple and adaptable method for estimating dog-population density and is not prone to problems associated with meeting some model assumptions inherent to mark-recapture estimators.     

家蚕产卵量的基因效应参数估计

用产卵量差异较大的两个近交系782和Ekp作亲本,配制了F1、F2正反交及F1与两亲本的回交B1和B2,加上两亲本P1和P2共8个世代,采用多世代平均值的多元回归分析方法对产卵量的基因效应参数进行了估计,得到了如下结果：（1）建立了分析家蚕产卵量的常染色体基因加性、显性和性染色体基因加性效应的非二体遗传模型,并通过实际数据检验表明产卵量遗传符合这个模式;（2）产卵量的遗传主要取决于常染色体上基因加性效应,其次是性染色体基因加性效应,显性效应很小。     

Small area estimations of proportion of forest and timber volume combining Lidar data and stereo aerial images with terrestrial data

Abstract

Methods for small area estimations were compared for estimating the proportion of forest and growing stock volume of temperate mixed forests within a district of a member state (canton) in Switzerland. The estimators combine terrestrial data with remotely sensed auxiliary data. By using different model types, different sources of auxiliary data and different methods of processing the auxiliary data, the increase in estimation precision was tested. Using the canopy height derived from remote sensing data, the growing stock volume and the proportion of forest were estimated. The regression models used for the small area estimation provided a coefficient of determination of up to 68% for the timber volume. The proportion of plots correctly classified into forest and non-forest plots ranged between 0.9 and 0.98. Models calibrated over forest area only resulted in a maximal coefficient of determination of 37%. Even though these coefficients indicate a moderate model quality, the use of remote sensing data clearly improved the estimation precision of both the proportion of forest and the growing stock volume. Generally, Lidar data led to slightly higher estimates compared to data from aerial photography. It was possible to reduce the variance of the estimated proportion of forest to nearly one tenth compared with the variance based on the terrestrial measurements alone. Similarly, the variance of the growing stock volume could be reduced to one fourth as compared with the variance based solely on the terrestrial measurements.     

变量筛选方法对郁闭度遥感估测模型的影响比较

比较基于偏最小二乘回归的Bootstrap方法与传统的平均残差平方和(RMSq)准则所选变量建立模型的精度差别.结果表明:Bootstrap方法是一种更优秀的变量筛选方法,比RMSq方法精度提高约5%;而且它不受变量多带来的运算困难的限制,更便于实际应用.