High-dimensional regression in practice: an empirical study of finite-sample prediction,variable selection and ranking

Penalized likelihood approaches are widely used for high-dimensional regression. Although many methods have been proposed and the associated theory is now well developed, the relative efficacy of different approaches in finite-sample settings, as encountered in practice, remains incompletely understood. There is therefore a need for empirical investigations in this area that can offer practical insight and guidance to users. In this paper, we present a large-scale comparison of penalized regression methods. We distinguish between three related goals: prediction, variable selection and variable ranking. Our results span more than 2300 data-generating scenarios, including both synthetic and semisynthetic data (real covariates and simulated responses), allowing us to systematically consider the influence of various factors (sample size, dimensionality, sparsity, signal strength and multicollinearity). We consider several widely used approaches (Lasso, Adaptive Lasso, Elastic Net, Ridge Regression, SCAD, the Dantzig Selector and Stability Selection). We find considerable variation in performance between methods. Our results support a “no panacea” view, with no unambiguous winner across all scenarios or goals, even in this restricted setting where all data align well with the assumptions underlying the methods. The study allows us to make some recommendations as to which approaches may be most (or least) suitable given the goal and some data characteristics. Our empirical results complement existing theory and provide a resource to compare methods across a range of scenarios and metrics.     

R&;D cooperation in emerging industries,asymmetric innovative capabilities and rationale for technology parks

Starting from the premise that firms are distinct in terms of their capacity to create innovations, this article explores the rationale for R&D cooperation and the choice between alliances that involve information sharing, cost sharing or both. Defining innovative capability as the probability of creating an innovation, it examines firm strategy in a duopoly market, where firms have to decide whether or not to cooperate to acquire a fixed cost R&D infrastructure that would endow each firm with a firm-specific innovative capability. Furthermore, since emerging industries are often characterized by high technological uncertainty and diverse firm focus that makes the exploitation of spillovers difficult, this article focuses on a zero spillover context. It demonstrates that asymmetry has an impact on alliance choice and social welfare, as a function of ex-post market competition and fixed costs of R&D. With significant asymmetry no alliance may be formed, while with similar firms the cost sharing alliance is dominant. Finally, it ascertains the settings under which the equilibrium outcome is distinct from that maximizing social welfare, thereby highlighting some conditions under which public investment in a technology park can be justified.     

Supporting pupils with special health needs in mainstream schools: policy and practice

Education policy favouring ‘inclusion’, together with medical advances, mean that a growing number of pupils in mainstream schools may have health‐related support needs in respect of a chronic illness or physical disability. Data from an empirical research study investigating these needs and carried out between 1996 and 1998 are used to reflect on the position of this group of pupils within policy guidance on special educational needs (SEN) and medical needs. Evidence of confusion and ambiguity, both in the guidance and its interpretation, suggests that the needs of this group remain somewhat hidden. More recent developments in special needs policy guidance are discussed in terms of the prospect for strengthening support for this group of pupils. Copyright © 2001 John Wiley & Sons, Ltd.     

Some partially sequential nonparametric tests for detecting linear trend

In the present study, we develop two nonparametric partially sequential tests for detecting possible presence of linear trend among the incoming series of observations. We assume that a sample of fixed size is available a priori from some unknown univariate continuous population and there is no sign of trend among these historical observations. Our proposed tests can be viewed as the sequential type tests for monitoring structural changes. We use partial sequential sampling schemes based on usual ranks as well as on sequential ranks. We provide detailed discussion on asymptotic studies related to the proposed tests. We compare the two tests under various situations. We also present some numerical results based on simulation studies. Proposed tests are extremely important in profit making in volatile market through Margin Trading. We illustrate the mechanism with a detailed analysis of a stock price data.     

Some inferential procedures in randomized repeated measurement design for binary response

We consider a two-treatment two-period crossover design in the presence of possible carryover effects, where the treatment responses are binary. We provide some simple probability models incorporating the possible carryover effects. Asymptotic distributions of the estimates of the parameters under the proposed models are derived. We carry out tests for treatment difference and carryover effects. Finally we use a data set to illustrate the applicability of the proposed procedures.     

Multi-arm response-adaptive designs for circular responses

A multi-arm response-adaptive allocation design is developed for circular treatment outcomes. Several exact and asymptotic properties of the design are studied. Stage-wise treatment selection procedures based on the proposed response-adaptive design are also suggested to exclude the worse performing treatment(s) at earlier stages. Detailed simulation study is carried out to evaluate the proposed selection procedures. The applicability of the proposed methodologies is illustrated through a real clinical trial data on cataract surgery.     

Antiracism Limited

Liberal enthusiasm over the arrival of a ‘post-racial’ era in the USA found ecstatic release at the 2008 election of the first African-American to the presidency. Several years have passed since that event, and the idea of ‘post-race’ remains a potent, if uneasy, keyword within the national imaginary. Authorizing ‘official antiracisms’ of our day, post-race sanctions the governmentalities of racial neoliberalism, and forecloses alternative paradigms for antiracist struggle. Equally, it registers as a cruel joke of sorts, turning on studied denials of abiding inequalities and open to the lash of ridicule and scorn. At once absurd and alluring, post-race reveals the cultural career of a keyword marked by power and authority as well as tenacious counter-claims that show up its conceits and specious allure.

A rich body of scholarship has emerged in recent years to unpack the meanings and implications of the term. These works typically take the Obama moment, unprecedented and consequential, as a key signifying theatre. This essay, however, begins elsewhere, suggesting instead that the 2008 victory constitutes a denouement, a narrative climax to epistemic struggles that shaped racial discourse over decades prior to and, indeed, anticipating Obama's arrival as post-racial icon. Offering a genealogical pre-history of post-race, the analysis uses the analytical lens of ‘bricolage’ to trace a range of articulations over a 40-year span that concludes with Obama's 2008 victory. The essay maps the riotous ascendance of post-race as well as its faltering claims to authority to reveal the conditions of possibility that gave rise to, and continue to provide traction for, the idea of the post-racial. Highlighting key epistemological shifts ushered in by the concept, the essay substantiates the terms of a third ‘racial break’ that works to foreclose all but those antiracist projects that serve the incentives of neoliberal capital.     

On compounded geometric distributions and their applications

Here, we introduce two-parameter compounded geometric distributions with monotone failure rates. These distributions are derived by compounding geometric distribution and zero-truncated Poisson distribution. Some statistical and reliability properties of the distributions are investigated. Parameters of the proposed distributions are estimated by the maximum likelihood method as well as through the minimum distance method of estimation. Performance of the estimates by both the methods of estimation is compared based on Monte Carlo simulations. An illustration with Air Crash casualties demonstrates that the distributions can be considered as a suitable model under several real situations.     

Fair and group strategy-proof good allocation with money

We completely characterize the class of fair and group strategy-proof mechanisms. We consider two notions of fairness, anonymity in welfare and no-envy. Both fairness axioms, when applied with strategy-proofness, imply decision efficiency, and lead to the same class of group strategy-proof mechanisms (where the group size is restricted to two). We find that the only feasible mechanism satisfying a mild zero transfer axiom, in this class, is the Pivotal mechanism.