Smoothing estimation of rate function for recurrent event data with informative censoring

This paper proposes kernel estimation of the occurrence rate function for recurrent event data with informative censoring. An informative censoring model is considered with assumptions made on the joint distribution of the recurrent event process and the censoring time without modeling the censoring distribution. Under the validity of the informative censoring model, we also show that an estimator based on the assumption of independent censoring becomes inappropriate and is generally asymptotically biased. To investigate the asymptotic properties of the proposed estimator, the explicit form of its asymptotic mean squared risk and the asymptotic normality are derived. Meanwhile, the empirical consistent smoothing estimator for the variance function of the estimator is suggested. The performance of the estimators are also studied through Monte Carlo simulations. An epidemiological example of intravenous drug user data is used to show the influence of informative censoring in the estimation of the occurrence rate functions for inpatient cares over time.     

Rhythmic growth of ring‐banded spherulites in blends of liquid crystalline methoxy‐poly(aryl ether ketone) and poly(aryl ether ether ketone)

Rhythmic growth of ring‐banded spherulites in blends of liquid crystalline methoxy‐poly(aryl ether ketone) (M‐PAEK) and poly(aryl ether ether ketone) (PEEK) has been investigated by means of differential scanning calorimetry (DSC), polarized light microscopy (PLM), and scanning electron microscopy (SEM) techniques. The measurements reveal that the formation of the rhythmically grown ring‐banded spherulites in the M‐PAEK/PEEK blends is strongly dependent on the blend composition. In the M‐PAEK‐rich blends, upon cooling, an unusual ring‐banded spherulite is formed, which is ascribed to structural discontinuity caused by a rhythmic radial growth. For the 50:50 M‐PAEK/PEEK blend, ring‐banded spherulites and individual PEEK spherulites coexist in the system. In the blends with PEEK as the predominant component, M‐PAEK is rejected into the boundary of PEEK spherulites. The cooling rate and crystallization temperature have great effect on the phase behavior, especially the ring‐banded spherulite formation in the blends. In addition, the effects of M‐PAEK phase transition rate and phase separation rate on banded spherulite formation is discussed. © 2007 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 45: 3011–3024, 2007     

基于进化人工神经网络的组合系统滤波技术

在组合系统运用Kalman滤波器技术时,准确的系统模型和可靠的观测数据是保证其性能的重要因素,否则将大大降低Kalman滤波器的估计精度,甚至导致滤波器发散.为解决上述Kalman应用中的实际问题,提出了一种新颖的基于进化人工神经网络技术的自适应Kalman滤波器.仿真试验表明该算法可以在系统模型不准确时、甚至外部观测数据短暂中断时,仍能保证Kalman滤波器的性能.     

光突发交换的交换控制策略和光缓存配置

光突发交换是面向下一代互联网的光交换模式.讨论了异步光突发交换系统的交换控制策略以及相应光缓存的优化配置策略.除了传统预约模式的交换策略,还研究了非预约和改进型预约模式,它们的性能评估由计算机仿真给出.结果表明:在条件相当情况下,改进型预约模式具有最低的丢包率.同时光缓存的配置对上述几类交换控制策略的性能都有很大影响,仿真结果指出:光缓存的粒度对系统性能具有重要影响,所讨论的几类交换模式都存在最佳时延粒度,研究结果对光交换矩阵的设计有指导意义.     

带有回报计划的动态客户关系管理模型及实验应用分析

在客户最大化效用及公司最大化CLV的动态环境下。对所提的带有回报计划的动态客户关系管理模型用于某超市的客户数据库中，发现模型的结果对这类客户是适用的。并给出了不同的客户状态空间对应的有效营销组合策略。结果表明：合适的回报计划可以促进客户的购买、提高公司的利润及缓解价格竞争。回报极限应该比客户的平均购买水平偏高，回报率应该与回报极限的改变方向一致。计划的时间范围应定在一年左右比较合适。对于累积购买水平较高的客户一般不邮寄商品信息。在回报计划的初期与末期不用打折。中期对那些购买次数很少的客户可以实行相应的降价策略。     

Miscibility and phase structure of binary blends of polylactide and poly(methyl methacrylate)

Blends of amorphous poly(DL‐lactide) (DL‐PLA) and crystalline poly(L‐lactide) (PLLA) with poly(methyl methacrylate) (PMMA) were prepared by both solution/precipitation and solution‐casting film methods. The miscibility, crystallization behavior, and component interaction of these blends were examined by differential scanning calorimetry. Only one glass‐transition temperature (T_g) was found in the DL‐PLA/PMMA solution/precipitation blends, indicating miscibility in this system. Two isolated T_g's appeared in the DL‐PLA/PMMA solution‐casting film blends, suggesting two segregated phases in the blend system, but evidence showed that two components were partially miscible. In the PLLA/PMMA blend, the crystallization of PLLA was greatly restricted by amorphous PMMA. Once the thermal history of the blend was destroyed, PLLA and PMMA were miscible. The T_g composition relationship for both DL‐PLA/PMMA and PLLA/PMMA miscible systems obeyed the Gordon–Taylor equation. Experiment results indicated that there is no more favorable trend of DL‐PLA to form miscible blends with PMMA than PLLA when PLLA is in the amorphous state. © 2002 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 41: 23–30, 2003     

On Non-Continuous Dirichlet Processes

We introduce here some Itô calculus for non-continuous Dirichlet processes. Such calculus extends what was known for continuous Dirichlet processes or for semimartingales. In particular we prove that non-continuous Dirichlet processes are stable under C ¹ transformation.     

基于CSCW的协同编著系统的设计

讨论了计算机支持的协同工作(CSCW)的主要技术，其中重点讨论了协同编著系统设计中的并发控制技术，指出了其与传统的并发控制技术的区别。根据CSCW的基本原理设计出了一个基于Browse／Server结构的协同编著系统的系统模型。该系统模型比传统的模型更加方便了用户之间通过Internet的浏览器进行协作。     

Effects of diamines and their fluorinated groups on the color lightness and preparation of organosoluble aromatic polyimides from 2,2‐bis[4‐(4‐amino‐2‐trifluoromethylphenoxy)phenyl]‐hexafluoropropane

To investigate the position and amount of the CF₃ group affecting the coloration of polyimides (PIs), we prepared 2,2‐bis[4‐(4‐amino‐2‐trifluoromethylphenoxy)phenyl]hexafluoropropane ( 2 ) with four CF₃ groups with 2‐chloro‐5‐nitrobenzotrifluoride and 2,2‐bis(4‐hydroxyphenol)hexafluoropropane. A series of soluble and light‐colored fluorinated PIs ( 5 ) were synthesized from 2 and various aromatic dianhydrides ( 3a – 3f ). 5a – 5f had inherent viscosities ranging from 0.80 to 1.19 dL/g and were soluble in amide polar solvents and even in less polar solvents. The glass‐transition temperatures of 5 were 221–265 °C, and the 10% weight‐loss temperatures were above 493 °C. Their films had cutoff wavelengths between 343 and 390 nm, b* values (a yellowness index) ranging from 5 to 41, dielectric constants of 2.68–3.01 (1 MHz), and moisture absorptions of 0.03–0.29 wt %. In a comparison of the PI series 6 – 8 based on 2,2‐bis[4‐(4‐aminophenoxy)phenyl]hexafluoropropane, 2,2‐bis[4‐(4‐amino‐2‐trifluoromethylphenoxy)phenyl]propane, and 2,2‐bis[4‐(4‐aminophenoxy)phenyl]propane, we found that the CF₃ group close to the imide group was more effective in lowering the color; this means that CF₃ of 5 , 7 , and 8f was more effective than that of 6c . The color intensity of the four PI series was lowered in the following order: 5 > 7 > 6 > 8 . The PI 5f , synthesized from diamine 2 and 4,4′‐hexafluoroisopropylidenediphthalic anhydride, had six CF₃ groups in a repeated segment, so it exhibited the lightest color among the four series. © 2003 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 41: 922–938, 2003     

基于DS/AHP的供应商选择方法

供应商选择方法有很多种，在众多的方法中层次分析法以能够将定性指标定量化而被广泛应用于供应商选择决策中。考虑到供应商选择问题中包含有很多的不确定性而证据理论在处理不确定问题又有着独特的优点，因此本文采用了一种由层次分析法和证据理论结合而产生的DS／AHP决策方法，并将其应用于供应商选择决策问题中，该方法综合了层次分析法和证据理论的优点，可以更科学的进行供应商选择决策，最后通过一个例子说明这种方法在供应商选择中的应用。