Non-linear regression models for Approximate Bayesian Computation

Approximate Bayesian inference on the basis of summary statistics is well-suited to complex problems for which the likelihood is either mathematically or computationally intractable. However the methods that use rejection suffer from the curse of dimensionality when the number of summary statistics is increased. Here we propose a machine-learning approach to the estimation of the posterior density by introducing two innovations. The new method fits a nonlinear conditional heteroscedastic regression of the parameter on the summary statistics, and then adaptively improves estimation using importance sampling. The new algorithm is compared to the state-of-the-art approximate Bayesian methods, and achieves considerable reduction of the computational burden in two examples of inference in statistical genetics and in a queueing model.     

Asymptotic bias in the linear mixed effects model under non-ignorable missing data mechanisms

Summary.  In longitudinal studies, missingness of data is often an unavoidable problem. Estimators from the linear mixed effects model assume that missing data are missing at random. However, estimators are biased when this assumption is not met. In the paper, theoretical results for the asymptotic bias are established under non-ignorable drop-out, drop-in and other missing data patterns. The asymptotic bias is large when the drop-out subjects have only one or no observation, especially for slope-related parameters of the linear mixed effects model. In the drop-in case, intercept-related parameter estimators show substantial asymptotic bias when subjects enter late in the study. Eight other missing data patterns are considered and these produce asymptotic biases of a variety of magnitudes.     

Saddlepoint approximations for some models of circular data

In this article we provide saddlepoint approximations for some important models of circular data. The particularity of these saddlepoint approximations is that they do not require solving the saddlepoint equation iteratively, so their evaluation is immediate. We first give very accurate approximations to P-values, critical values and power functions for some optimal tests regarding the concentration parameter under wrapped symmetric α-stable and circular normal models. Then, we consider an approximation to the distribution of a projection of the two-dimensional Pearson random walk with exponential step sizes.     

Discovering influential cases in linear regression with MINITAB: Peeking into multidimensions with a MINITAB macro

Various diagnostic statistics have been proposed to help identify cases that markedly affect, or influence, the features of a fitted linear regression model. Once influential cases are found, decisions can be made regarding their worth in the model building process. Since a subject data set may contain both singly influential cases and influential multiple case subsets, the capability to assess the joint influence of cases is needed for a complete analysis. The aim of this work is to briefly review Cook’s distance measure for multiple cases, an effective diagnostic for this purpose, and present a method using it to search for influential multiple case subsets. The method is applied in two example analyses by way of a MINITAB Statistical Software macro.     

计算机与学术写作(英文)

深大外语系把若干计算机软件应用于英语专业本科学生的写作课教材系列。这不仅有助于学生自修写作技巧,且能为教师节省大量时间,使教师有可能对学生作更多的个别辅导。同时,学生在上写作课的过程中还能掌握一种使用计算机的实际技能。近十年来,中国的飞速发展是史无前例的。西方的技术发展史对中国虽然有启迪作用,但其实际要求与中国相去甚远。在这个意义上,深大的尝试对全中国均具有示范意义。     

Quantifying Economic Dependency

In this paper we compare several types of economic dependency ratios for a selection of European countries. These dependency ratios take into account not only the demographic structure of the population, but also the differences in age-specific economic behaviour such as labour market activity, income and consumption as well as age-specific public transfers. In selected simulations where we combine patterns of age-specific economic behaviour and transfers with population projections, we show that in all countries population ageing would lead to a pronounced increase in dependency ratios if present age-specific patterns were not to change. Our analysis of cross-country differences in economic dependency demonstrates that these differences are driven by both differences in age-specific economic behaviour and in the age composition of the populations. The choice of which dependency ratio to use in a specific policy context is determined by the nature of the question to be answered. The comparison of our various dependency ratios across countries gives insights into which strategies might be effective in mitigating the expected increase in economic dependency due to demographic change.     

Organizational ecology of immigrant employment and organizational buffer zones: Who leaves first when the organization gets into trouble?

Using data from the German social insurance system immigrant employment in organizational buffer zones will be investigated from an organizational ecology perspective. Theories in organization and labor market research predict higher exit rates of employees holding jobs in buffer zones. These buffer zones protect organizations from unfavorable environmental turbulences. Consequently, analyzing the impact of buffer zone employment on job stability of immigrants requires a fully dynamic approach, which goes beyond a mere identification of generally increased turnover rates of immigrants.In the first step, determinants of organizational dissolution rates will be estimated. In order to obtain a dynamic measurement of unfavorable environmental conditions, the model includes both organizational and time-varying environmental characteristics of organizational populations. This will be done by using a three-level mixed-effects hazard model controlling for observed and unobserved environmental characteristics. In the second step, dissolution rates predicted in the first step will be used as an explanatory variable. These rates indicate the degree of enactment of the flexible buffer zone. It will be shown that immigrants have indeed higher turnover rates, but they tend to remain longer in their organizations when environmental conditions deteriorate—and they are not those who leave first when the organization gets into trouble.     

The differential contributions of teen drinking homophily to new and existing friendships: An empirical assessment of assortative and proximity selection mechanisms

Alcohol use is pervasive in adolescence. Though most research is concerned with how friends influence drinking, alcohol is also important for connecting teens to one another. Prior studies have not distinguished between new friendship creation, and existing friendship durability, however. We argue that accounting for distinctions in creation–durability processes is critical for understanding the selection mechanisms drawing drinkers into homophilous friendships, and the social integration that results. In order to address these issues, we applied stochastic actor based models of network dynamics to National Longitudinal Study of Adolescent Health data. Adolescents only modestly prefer new friendships with others who drinker similarly, but greatly prefer friends who indirectly connect them to homophilous drinkers. These indirect homophilous drinker relationships are shorter lived, however, and suggest that drinking is a social focus that connects adolescents via proximity, rather than assortativity. These findings suggest that drinking leads to more situational and superficial social integration.