Hamiltonian Monte Carlo acceleration using surrogate functions with random bases

For big data analysis, high computational cost for Bayesian methods often limits their applications in practice. In recent years, there have been many attempts to improve computational efficiency of Bayesian inference. Here we propose an efficient and scalable computational technique for a state-of-the-art Markov chain Monte Carlo methods, namely, Hamiltonian Monte Carlo. The key idea is to explore and exploit the structure and regularity in parameter space for the underlying probabilistic model to construct an effective approximation of its geometric properties. To this end, we build a surrogate function to approximate the target distribution using properly chosen random bases and an efficient optimization process. The resulting method provides a flexible, scalable, and efficient sampling algorithm, which converges to the correct target distribution. We show that by choosing the basis functions and optimization process differently, our method can be related to other approaches for the construction of surrogate functions such as generalized additive models or Gaussian process models. Experiments based on simulated and real data show that our approach leads to substantially more efficient sampling algorithms compared to existing state-of-the-art methods.     

Racial Disparities in Welfare Dependence and Financial Independence: Links to Human Capital,Local Economy,and State Temporary Assistance for Needy Families Policies

This secondary data analysis examined racial disparities in associations betwen welfare dependence/financial independence and human capital, local economy, and state TANF policies. A sample of 6,737 parents was extracted from the public-use data set titled “National Longitudinal Survey of Youth.” Results showed that restrictive TANF policies reduced African Americans’ likelihood of welfare use and increased likelihood of their financial independence. Multinomial logistic results also showed that, among Hispanics, employment growth in neighboring counties promoted welfare use; whereas among Caucasians such growth promoted financial independence. County poverty increased (a) Caucasians’ likelihood of welfare use and (b) Hispanics’ likelihood of being working poor; it decreased Caucasians’ and African Americans’ likelihood of financial independence. Across ethnic groups, education reduced likelihood of welfare use and working poor status; across minority groups, education increased likelihood of financial independence, but among Caucasians it decreased such likelihood. Across ethnic groups, occupational skills hindered dependence and improved odds of employment (regardless of welfare or poverty status). This study concluded the studied TANF policies and job markets were not color-blind. Interventions this study implies include less-restrictive TANF policies, generous support services, TANF staff cultural-competence training, and antidiscrimination rules. Research investigating particular TANF policies’ and services’ effects by ethnicity might prove useful.     

广西少数民族非物质文化遗产数字化保护现状及对策分析

广西作为少数民族最为聚集的地区之一,拥有极为丰富的民族文化资源.数字化是推进少数民族地区非物质文化遗产保护和传承的重要手段.本文从体制机制、人才资金、数字化手段、新媒体传播四个角度出发,对广西少数民族非物质文化遗产的数字化现状进行调研,分析其存在的问题,并从如何树立正确的保护理念、完善数字化体制机制、实施多样性的数字化保护手段等五个方面提出对策建议,以推进广西少数民族非物质文化遗产数字化的可持续发展.     

两变量异方差随机系数回归模型的最优设计

文章研究了具有异方差误差结构的两变量随机系数回归模型的最优设计,证明了当异方差误差结构满足一个充分条件时,模型在几类经典设计准则下的最优设计可以在设计域的顶点处获得.     

Instrumental variable estimation in ordinal probit models with mismeasured predictors

Researchers in the medical, health, and social sciences routinely encounter ordinal variables such as self‐reports of health or happiness. When modelling ordinal outcome variables, it is common to have covariates, for example, attitudes, family income, retrospective variables, measured with error. As is well known, ignoring even random error in covariates can bias coefficients and hence prejudice the estimates of effects. We propose an instrumental variable approach to the estimation of a probit model with an ordinal response and mismeasured predictor variables. We obtain likelihood‐based and method of moments estimators that are consistent and asymptotically normally distributed under general conditions. These estimators are easy to compute, perform well and are robust against the normality assumption for the measurement errors in our simulation studies. The proposed method is applied to both simulated and real data. The Canadian Journal of Statistics 47: 653–667; 2019 © 2019 Statistical Society of Canada     

Objectifying or Liberating? Investigation of the Effects of Sexting on Body Image

Scholars are divided as to whether sexting—an unprecedented sexual activity using digital media—is objectifying or sexually liberating. One notion is that sexting involves the representation of an individual’s sexuality in the presence of others and thus reinforces objectification. Another perspective contends that the self-portrayal of the body in sexting facilitates the exploration of sexual subjectivity and is, therefore, sexually liberating. By testing a model of sexting, objectified body consciousness (body surveillance, body shame, and body control beliefs), and comfort with nudity (indicator of sexual liberation) on 361 college students in Hong Kong, the current study revealed that, across genders, sexters demonstrated higher levels of body surveillance, body shame, and comfort with nudity than nonsexters. The results suggest that sexting is both sexually objectifying and liberating and that it has opened up a new sexual arena that combines sexual objectification and empowerment.     

工业生产均衡性统计之我见

目前对工业生产均衡性统计还很不完善。主要表现在两个方面:一是生产均衡性的含义模糊不清;二是生产均衡率的计算不科学。对于均衡率的计算,统计界有不少学者进行过研究,但对于均衡性的含义却很少有人问津。为了完善工业生产均衡性统计,本文就这两方面作了一些初步的探索,不当之处难免,望统计界的同行们予以指正。     

An Empirical Model for Managing Quality in the Electronics Industry

Much of the empirical research in the past two decades has suggested that quality management (QM) is context dependent. This research develops an empirical QM model in a technology‐based sector—electronics manufacturing. Based on quantitative and qualitative investigations of 225 electronics firms in Hong Kong and the Pearl River Delta (PRD) region of China, a path analytic model is developed. The empirical model shows that a typical quality management system (QMS) in the electronics industry is composed of four major modules, namely leadership, cultural elements, operational support systems, and process management. These modules create a series of chain effects on organizational performance, rather than acting as parallel elements with an equal impact. By quantifying their effects on organizational performance and comparing the model to others in the literature, we identify those QM constructs that are context dependent. In electronics manufacturing, process management and customer focus are more important than other elements (e.g., cultural factors) for garnering business results. This study contributes to contingency theory and research by identifying the key constructs and their relationships in a competitive, volatile, and technology‐based industry with complex supply networks.