On partial linear additive isotonic regression

In this paper we discuss semiparametric additive isotonic regression models. We discuss the efficiency bound of the model and the least squares estimator under this model. We show that the ordinary least square estimator studied by Huang (2002) and Cheng (2009) for the semiparametric isotonic regression achieves the efficiency bound for the regular estimator when the true parameter belongs to the interior of the parameter space. We also show that the result by Cheng (2009) can be generalized to the case that the covariates are dependent on each other.     

Optimal hedging through limit orders

This article applies the methods of stochastic dynamic programming to a risk management problem, where an agent hedges her derivative position by submitting limit orders. Therefore, this model is the first, in the literature on optimal trading with limit orders, to handle a problem of hedging options or other derivatives. A hedging strategy is developed where both the size and the limit price of each order is optimally set.     

Complex Latent Variable Modeling in Educational Assessment

Bayesian item response theory models have been widely used in different research fields. They support measuring constructs and modeling relationships between constructs, while accounting for complex test situations (e.g., complex sampling designs, missing data, heterogenous population). Advantages of this flexible modeling framework together with powerful simulation-based estimation techniques are discussed. Furthermore, it is shown how the Bayes factor can be used to test relevant hypotheses in assessment using the College Basic Academic Subjects Examination (CBASE) data.     

Local estimation for varying-coefficient models with longitudinal data

Varying-coefficient models are very useful for longitudinal data analysis. In this paper, we focus on varying-coefficient models for longitudinal data. We develop a new estimation procedure using Cholesky decomposition and profile least squares techniques. Asymptotic normality for the proposed estimators of varying-coefficient functions has been established. Monte Carlo simulation studies show excellent finite-sample performance. We illustrate our methods with a real data example.     

Zen and the art of aut-ethnography: a tribute to Robert M. Pirsig

With the passing of Robert M. Pirsig, I felt that it was an appropriate time to write a tribute to his work and the influence it has had on my own theorising in regard to autistic ways of being. This reflection utilises the concept of an ‘aut-ethnography’ to examine passages that I had highlighted word by word when I first read Pirsig’s book: Zen and the Art of Motorcycle Maintenance. These fragments contain links to a number of theoretical ‘lines of flight’ within my own work and that of others, from his concepts of dynamic ‘quality’ to his discussion on the tension between scientific method and lived experience.     

Attributing Outcomes to Social Policy Interventions – ‘Gold Standard’ or ‘Fool's Gold’ in Public Policy and Management?

In recent years there has been increased interest in outcome‐based social policy‐making and management. The UK has been in the forefront of this movement but similar movements have been identified internationally. This interest in outcome‐based decision‐making has been given particular impetus through the ‘results’‐based movement in evaluation and performance management since the 1980s, which has increased in scope over time, slowly changing its emphasis from cost reduction and measuring outputs to measuring outcomes. This change has been widely welcomed by policymakers, practitioners and academics. However, there is evidence that the reality is often rather less than the rhetoric. Moreover, the ‘attribution problem’ of attributing changes in outcomes to specific social policies has remained a major issue. The conceptual solution of constructing ‘cause‐and‐effect’ models, imported from the policy evaluation field, has only recently become common for operationalising these models. This article outlines the evolution of interest in outcome‐based social policy‐making up to recent times and the growing realization of the importance of the attribution problem. It then outlines both how the ‘cause‐and‐effect’ policy modelling approach can partially tackle the attribution problem, but also its inherent limitations. Lastly, the article uses several case studies in current UK social policy‐making to demonstrate the potential importance of the reasoning embedded within cause‐and‐effect models but also the dangers in policy‐making which adopts this approach without understanding its conceptual basis or in fields where it is inappropriate, given the current state of our knowledge of social policy systems.     

人力资源管理与企业战略、绩效关系研究的演进——基于企业环境的变化

高度动态化的环境作为当前企业面临的最大问题,导致不可预测性的增强。这将直接影响战略的制定和执行。因此,这一环境下的企业如何保持战略实施有效性和竞争优势可持续性成为企业发展道路上的关键。而在战略与绩效的关系中,人力资源作为企业关键的战略性资源是企业运作的基础,其管理起了非常重要的作用。本文通过对已有文献的详细梳理,概括出两类研究类型:战略和人力资源的匹配;人力资源的配置管理与实现绩效的关系。柔性化思想的提出,使人力资源研究有了新的方向。在环境变动这一前提下,人力资源柔性化的管理区别于传统意义上的管理。本文以动态环境的背景,围绕人力资源、战略、绩效三者的关系提炼出整合的演进框架,以更好地指导未来人力资源相关的研究。     

Statistical estimation of internal available bandwidth

This paper presents a method for using end-to-end available bandwidth measurements in order to estimate available bandwidth on individual internal links. The basic approach is to use a power transform on the observed end-to-end measurements, model the result as a mixture of spatially correlated exponential random variables, carryout estimation by moment methods, then transform back to the original variables to get estimates and confidence intervals for the expected available bandwidth on each link. Because spatial dependence leads to certain parameter confounding, only upper bounds can be found reliably. Simulations with ns2 show that the method can work well and that the assumptions are approximately valid in the examples.     

Bivariate symbolic regression models for interval-valued variables

Interval-valued variables have become very common in data analysis. Up until now, symbolic regression mostly approaches this type of data from an optimization point of view, considering neither the probabilistic aspects of the models nor the nonlinear relationships between the interval response and the interval predictors. In this article, we formulate interval-valued variables as bivariate random vectors and introduce the bivariate symbolic regression model based on the generalized linear models theory which provides much-needed exibility in practice. Important inferential aspects are investigated. Applications to synthetic and real data illustrate the usefulness of the proposed approach.