异质性大数据的分布式估计

随着物联网技术的进步,大数据给网络带宽和计算机存储能力带来巨大挑战,传统的集中式数据处理难以实现,客观上促进了分布式统计学习的发展。在无迭代算法研究中,Zhang等(2013)证明了当数据集个数s=O(N) 时,基于局部经验风险最小化的分治(DC)简单平均估计量具有O(N-1)均方误差收敛速度,Huang和Huo(2019)在M估计框架下进一步提出分布式一步估计量,但上述方法均未考虑海量数据可能存在的异质性对分治估计效果的影响。本文在线性模型框架下提出海量异质数据的分治一步加权估计,证明了估计量的渐近性质并考虑了异质性检验问题。将本文提出的方法应用于美国医疗保险实际数据分析,结果表明该方法能更好地拟合数据的线性趋势且显著提高了计算效率。     

大规模数据下基于充分降维的Leverage重要性抽样方法

随着信息技术的飞速发展，大规模数据在短时间内搜集并储存下来，为分析决策提供了巨大的信息量，也给统计建模带来了一定难度。对于样本容量大、变量个数少的数据，Leverage重要性抽样是一个简便可行的方法。本文发现，该方法中度量样本重要性的Leverage分数与因变量无关，而且在维度较大的情形下对样本没有区分程度，使得估计结果较差。为了同时考虑因变量和维度的影响，本文提出了基于充分降维的Leverage重要性抽样方法。该方法以不损失信息为前提，在充分降维的空间内重新计算Leverage分数，使得抽样更具有代表性。模拟数据分析显示，在样本容量较大的复杂数据中，相比于原始的Leverage重要性抽样方法，本文提出的方法可以降低估计的均方误差。三个实际数据也证实了该方法的可行性和有效性。     

Rapid Gender Analysis and its use in crises: from zero to fifty in five years

ABSTRACT

This article explores how and why the international non-government organisation, CARE, developed its own system of gender analysis, Rapid Gender Analysis (RGA), during the humanitarian response in Syria. The article tracks and reviews a sample of the first 50 CARE RGA reports to share recurrent gender themes that emerge across them, including the lack of women’s meaningful participation in decision-making, limitations on women and girls’ mobility, increased risks of gender-based violence, and recurring issues facing humanitarian organisations in providing a gender-sensitive response. RGA has now been used in more than 50 crises around the world and is featured as a good practice in the Inter-Agency Standing Committee Gender Handbook. It is giving humanitarians faster and more complete access to information about gender norms than ever before. But, this article asks, has the RGA made a difference and, if so, to whom?     

Efficiency and Productivity changes in the presence of undesirable outcomes in Emergency Obstetric and Newborn Care

This paper analyzes efficiency and productivity changes in Emergency Obstetric and Newborn care (EmONC) in regional public hospitals in India. A novel approach of Weighted Russell Directional Distance Function Model (WRDDFM) has been employed. This model considers all the radial and non-radial slacks associated with good outputs as well as the undesirable outcomes while estimating inefficiency and helps to identify the source of the inefficiency. The empirical results based on a sample of 46 regional public hospitals from 2006–07 to 2013–14 show on average overall inefficiency of 24% and much of these are due to inefficiencies associated with undesirable outcome of stillbirths and intra-uterine deaths. Further, on an average the sampled hospitals have experienced marginal productivity decline over the reform period. The study indicates that there is a greater scope for efficiency and productivity gains in public hospitals, particularly if the undesirable outcome of still birth is minimized.     

通风机性能的微机全自动测试与数据处理系统

介绍一种通风机性能微机全自动测试与数据处理系统,阐述了该系统的组成和功能,传感器和主要设备的选用,系统软件的设计及提高系统测试精度所采用的方法。工业应用证实该系统性能良好。     

Some subjective bayesian considerations in the selection of models

Some considerations relating to the post-data selection of models are discussed. These include some difficulties with orthodox theory, implementation of the likelihood principle, and Bayesian tests of hypotheses.     

一种用于人发中微量元素铬测定的新消化方法

本文用原子吸收分光光度法测试人发中微量元素铬的含量,用1:1的硝酸,在高压消化罐中消化处理发样,测试结果,比之用其他化学方法消化发样,测试精度高,回收率好,一般在96%～107%.且节省试剂,操作简单,是目前较理想的样品消化方法.     

仪器分析与有机化学共同开设综合性实验的尝试

该文介绍了林业院校少学时的有机化学实验和仪器分析实验共同开设综合性实验的尝试。将有机化学实验课上制备的产品,在仪器分析实验中进行测试分析,使实验教学更接近科学研究和生产实际,既节省了学时,又提高了学生对实验课的兴趣,有利于培养学生的科研能力和综合实验素质。