A comparison of various methods for estimating the parameters in mixtures of von mises distributions

In this paper we compare five methods for estimating the unknown parameters in a mixture of two von Mises distributions. We propose a new method based on the characteristic function and compare it with the classical methods based on maximum likelihood and moments. Thus far these methods have been successfully applied only to linear data. Here we show that the application to circular data is reasonably straightforward and that convergence to the final estimates is fairly rapid. For various simulated known mixtures the results obtained are satisfactory. Finally, we introduce a modification of the method of moments which is considerably faster in CPU time than any of the other methods used and gives good results.     

Time Series Modeling of Degradation Due to Outdoor Weathering

Environmental variables have an important effect on the reliability of many products such as coatings and polymeric composites. Long-term prediction of the performance or service life of such products must take into account the probabilistic/stochastic nature of the outdoor weather. In this article, we propose a time series modeling procedure to model the time series data of daily accumulated degradation. Daily accumulated degradation is the total amount of degradation accrued within one day and can be obtained by using a degradation rate model for the product and the weather data. The fitted model of the time series can then be used to estimate the future distribution of cumulative degradation over a period of time, and to compute reliability measures such as the probability of failure. The modeling technique and estimation method are illustrated using the degradation of a solar reflector material. We also provide a method to construct approximate confidence intervals for the probability of failure.     

Designs for estimating the difference between two responses

In this paper the problem of designing experiments for estimating the difference between responses at two arbitrary points in the region of interest is considered. Under the average mean squared error criterion the design moment conditions have been worked out. The results have been applied to the class of central composite designs.     

数学美在科学探索中的功能

本文通过对数学和自然科学发展史的考察,认为数学美在科学家的科学实践活动中,发挥着重要的动力作用和方法论功能。即数学美为科学探索提供持久的内在动力;为科学发现提供武器和导引;为科学成果提供选择和评价的标准。     

固定效应部分线性变系数面板模型的快速有效估计

本文将最小二乘支持向量机(LSSVM) 和二次推断函数法(QIF) 相结合，为个体内具有相关结构的固定效应部分线性变系数面板模型提供了一种新的快速估计方法；在一定的正则条件下，论证了参数估计量的渐近正态性和非参数估计量的收敛速度；采用Monte Carlo模拟考察了估计方法在有限样本下的表现并将估计技术应用于现实数据分析。该方法不仅保证了估计的有效性和统计推断力，而且程序运行速度得到较大幅度提升。     

“三合一”型政府管理模式:现实与理论的全景式展示

建设法治型政府、责任型政府和服务型政府是我国政府进行政府改革的有益尝试,是用新的价值取向和路径选择对传统政府管理模式进行根本性或方向性的调整.本文将对这三种政府管理模式作一番现实考量和理论分析,以求正确处理好它们之间的关系,努力彰显出它们在目前我国市场经济发展水平和社会整体发展水平中的本色和个性,并使其相辅相成、相互促进,塑造"三合一"型政府管理模式,可以实现社会理想的崇高意蕴,契合"和谐社会"的构建,这对促进我国经济社会协调发展和全面进步,具有重大的理论和实践意义.     

论高校学报质量函数关系

高等学校学报质量是一个整体概念,要在策划、选题、审稿、编校、装帧设计、印制、发行等各个环节都以质量为本,以求达到最佳质量效果。每一个环节的质量形成了一个质量体系,这个质量体系包含多个参数组成,每一个参数对高等学校学报质量作用不同,从而这些参数与学报整体质量形成函数关系。     

新疆回族驼运业的调查与研究

伴随着自元代起始的徙居,地处丝绸之路孔道的新疆回族便开始了远距离经商贸易活动。作为集货物运输与商业贸易为一体的商旅组织,回族驼运业随丝绸之路的兴衰而消长。到民国后,回族驼运业迅猛发展,在新疆驼运业中,规模最大,经营活动范围最广,资金最雄厚,社会影响最巨。本文依据田野调查资料及文献记载,对新疆回族驼运的组织与功能、通商路线与营运货物、驼运业兴起的原因及驼队的演变、新疆回族现在的经商运输情况等,作了系统论述。     

实施"兴边富民行动"政策的理论思考

本文从边疆政策的历史沿革、"兴边富民行动"的历史背景、政策取向等方面,深入探讨了"兴边富民行动"政策的作用和意义。     

A Stochastic Goal Program for Employee Scheduling*

Deterministic goal programs for employee scheduling decisions attempt to minimize expected operating costs by assigning the ideal number of employees to each feasible schedule. For each period in the planning horizon, managers must first determine the amount of labor that should be scheduled for duty. These requirements are often established with marginal analysis techniques, which use estimates for incremental labor costs and shortage expenses. Typically, each period in the planning horizon is evaluated as an independent epoch. An implicit assumption is that individual employees can be assigned to schedules with as little as a single period of work. If this assumption violates local work rules, the labor requirements parameters for the deterministic goal program may be suboptimal. As we show in this research, this well-known limitation can lead to costly staffing and scheduling errors. We propose an employee scheduling model that overcomes this limitation by integrating the labor requirements and scheduling decisions. Instead of a single, externally determined staffing goal for each period, the model uses a probability distribution for the quantity of labor required. The model is free to choose an appropriate staffing level for each period, eliminating the need for a separate goal-setting procedure. In most cases this results in better, less costly decisions. In addition, the proposed model easily accommodates both linear and nonlinear under- and overstaffing penalties. We use simple examples to demonstrate many of these advantages and to illustrate the key techniques necessary to implement our model. We also assess its performance in a study of more than 1,700 simulated stochastic employee scheduling problems.