A study on imbalance support vector machine algorithms for sufficient dimension reduction

Li et al. (2011 Li, B., Artemiou, A., Li, L. (2011). Principal support vector machine for linear and nonlinear sufficient dimension reduction. Ann. Stat. 39:3182–3210.[Crossref], [Web of Science ®] , [Google Scholar]) presented the novel idea of using support vector machines (SVMs) to perform sufficient dimension reduction. In this work, we investigate the potential improvement in recovering the dimension reduction subspace when one changes the SVM algorithm to treat imbalance based on several proposals in the machine learning literature. We find out that in most situations, treating the imbalanced nature of the slices will help improve the estimation. Our results are verified through simulation and real data applications.     

The education of children in care and children in need: Who falls behind and when?

We seek to explain the development of the educational gap between children in “out‐of‐home care” (CLA), children deemed in social need (CIN), and other pupils. A cohort of 642,805 pupils aged 16 in 2013 was used to chart the educational progress of the full cohort, the CLA (n = 6,236), the CIN in 2012 or 2013 but not CLA (n = 20,384), and a sample individually matched with the CLA (n = 11,084). At age 7, attainment of the CLA and CIN was approximately 1 standard deviation lower than the cohort average and predicted attainment at 16. At this point, the persistent “CIN” (those with earlier and persistent needs) had the lowest attainment relative to others, and this declined further during secondary school. Those entering care before or during primary school had very low attainment at age 7, but their relative attainment did not decline. Attainment of CLA and CIN at age 16 likely reflects early environment, special educational needs, and poor relationships with secondary school. Policy, research, and intervention should focus on CIN as well as CLA, do so before entry to care, and take account of the onset of, and probable reasons for, educational difficulties.     

A qualitative exploration of reentry service needs: The case of fathers returning from prison

The challenge of community and family reentry after a parent's release from prison remains an under‐addressed area of collateral damage stemming from high rates of U.S. incarceration. Many fathers released from prison return to living with family, and later attribute family connections and parent–child contact as key factors in their postrelease success. However, reentry planning is hampered by a dearth of research on family‐focused reentry services, and consequently, often omits attention to resuming family and parenting roles. To address this gap, we conducted 38 semi‐structured interviews with 19 previously incarcerated fathers, 9 co‐parenting mothers, and 10 relatives to explore service needs of fathers during reentry. Findings suggest programs not only should be multimodal, emphasizing family connections complemented by socioeconomic, self‐care, and social support services, but also should be accessible and relatable, offered within the community, and engaging for fathers and family members. Findings reinforce the importance of self‐determination and human agency while underscoring the multiple challenges fathers face upon reentry. By including the voices of those most affected by incarceration, this study advances knowledge to shape reentry programs and policies, contributes to efforts addressing criminal justice inequities, and promotes well‐being among formerly incarcerated parents and their families.     

Randomization Inference for Outcomes with Clumping at Zero

While randomization inference is well developed for continuous and binary outcomes, there has been comparatively little work for outcomes with nonnegative support and clumping at zero. Typically, outcomes of this type have been modeled using parametric models that impose strong distributional assumptions. This article proposes new randomization inference procedures for nonnegative outcomes with clumping at zero. Instead of making distributional assumptions, we propose various assumptions about the nature of the response to treatment and use permutation inference for both testing and estimation. This approach allows for some natural goodness-of-fit tests for model assessment, as well as flexibility in selecting test statistics sensitive to different potential alternatives. We illustrate our approach using two randomized trials, where job training interventions were designed to increase earnings of participants.     

Model‐free causal inference of binary experimental data

For binary experimental data, we discuss randomization‐based inferential procedures that do not need to invoke any modeling assumptions. In addition to the classical method of moments, we also introduce model‐free likelihood and Bayesian methods based solely on the physical randomization without any hypothetical super population assumptions about the potential outcomes. These estimators have some properties superior to moment‐based ones such as only giving estimates in regions of feasible support. Due to the lack of identification of the causal model, we also propose a sensitivity analysis approach that allows for the characterization of the impact of the association between the potential outcomes on statistical inference.     

The role of The Boffin as abject Other in gendered performances of school achievement

The gendered subjectivities of high achieving school pupils are examined, demonstrating the uneasy relationship between high educational achievement and peer popularity. Drawing on data from a study involving classroom observation and interviews with 71 high-achieving pupils across nine secondary schools in England, the article focuses on the gendered construction of The Boffin. Judith Butler's work, and Hannah Arendt's conception of pariahs and parvenus, are applied to analyse The Boffin as pariah, and as 'queer' in the classroom. The article explores the ways in which this construction is gendered, and the consequences of its application for girls and boys. It is argued that application of this stigmatised term functioned as a powerful deterrent to 'excessive' performances of achievement, for all pupils. But the risks are particularly acute for boys, to whom the 'naming' of Boffin appeared more readily applied, and for whom marginalization as Boffin appears especially perilous (risking particularly pernicious forms of punishment at the hands of other boys).     

A dynamic programming approach of finding an optimal broadcast schedule in minimizing total flow time

We study the problem of (off-line) broadcast scheduling in minimizing total flow time and propose a dynamic programming approach to compute an optimal broadcast schedule. Suppose the broadcast server has k pages and the last page request arrives at time n. The optimal schedule can be computed in O(k³(n+k)^k−1) time for the case that the server has a single broadcast channel. For m channels case, i.e., the server can broadcast m different pages at a time where m < k, the optimal schedule can be computed in O(n^k−m) time when k and m are constants. Note that this broadcast scheduling problem is NP-hard when k is a variable and will take O(n^k−m+1) time when k is fixed and m ≥ 1 with the straightforward implementation of the dynamic programming approach. The preliminary version of this paper appeared in Proceedings of the 11th Annual International Computing and Combinatorics Conference as “Off-line Algorithms for Minimizing the Total Flow Time in Broadcast Scheduling”.     

Informational impediments to the use of capital budgeting models

The primary purpose of this study was to empirically examine whether or not informational impediments impair or alter decision making, with specific focus on capital budgeting decisions. Based on detailed interviews of senior corporate executives, it appears that management information systems (MIS) are less than adequate in providing the information items required to utilize the non-naive methods of capital budgeting. It is therefore suggested that the MIS concept be vastly broadened in regard to such decisions.