Value-at-risk support vector machine: stability to outliers

A support vector machine (SVM) stable to data outliers is proposed in three closely related formulations, and relationships between those formulations are established. The SVM is based on the value-at-risk (VaR) measure, which discards a specified percentage of data viewed as outliers (extreme samples), and is referred to as \(\mathrm{VaR}\) -SVM. Computational experiments show that compared to the \(\nu \) -SVM, the VaR-SVM has a superior out-of-sample performance on datasets with outliers.     

Integrating resilience into floodplain restoration

This paper explores the role of geographic prioritization as a means of identifying lands that are especially well suited to restoration in the historical floodplain of western Oregon's Willamette River. The river and its floodplain have been extensively modified since 1850. As a result, the quantity and quality of river habitat have declined. The approach illustrated here explicitly links the potential for ecological benefits from restoration with the social and economic likelihood of restoration success. Using a consistent analytical framework, longitudinal patterns in selected biophysical and socioeconomic characteristics are quantified along the entire river length, from Eugene to Portland. Areas with high potential for ecological recovery and low socioeconomic constraints have the greatest potential for future restoration. Areas that combine low potential for ecological response with high demographic and economic costs are likely to be poor choices for restoration. Areas with high ecological potential and intermediate levels of socioeconomic constraints present intermediate opportunities for restoration. The paper identifies high priority locations for restoration, assuming the purposes of restoration are to increase river channel complexity, increase floodplain forest area, and increase natural water storage during floods. It concludes by discussing the role of visualizations in citizen involvement.     

Analytical closed-form solution for binary logit regression by categorical predictors

In contrast to the common belief that the logit model has no analytical presentation, it is possible to find such a solution in the case of categorical predictors. This paper shows that a binary logistic regression by categorical explanatory variables can be constructed in a closed-form solution. No special software and no iterative procedures of nonlinear estimation are needed to obtain a model with all its parameters and characteristics, including coefficients of regression, their standard errors and t-statistics, as well as the residual and null deviances. The derivation is performed for logistic models with one binary or categorical predictor, and several binary or categorical predictors. The analytical formulae can be used for arithmetical calculation of all the parameters of the logit regression. The explicit expressions for the characteristics of logit regression are convenient for the analysis and interpretation of the results of logistic modeling.     

A method from the ‘lifeworld’: some possibilities for person centred planning for children in care

This paper describes a method for working with children who are the subjects of care planning and review under the Children Act 1989. The person centred planning method, as it is termed, has been well established in working with adults with special needs but can be extrapolated to encounters with children. It focuses on three fundamental areas: relationship, meaning and narrative. In underscoring these areas, the method restrains the bureaucracy and experience of stigma that is often present for those residing in state care. On a wider plane, the paper argues that the method stands as a veritable example of Habermas's ideas concerning the application of communicative rationality and discursive democracy in social life.     

African American MSW Students: Personal Influences on Social Work Careers and Factors in Graduate School Selection

The research is an empirical examination of who the individuals were that influenced 207 African American respondents to pursue graduate social work education, and what enrollment decision factors (EDFs) led them to select a social work program. Statistically significant relationships between EDFs are also isolated. The purposive sample consisted of MSW graduates from three predominantly White universities in two southeastern states in the United States. Almost 80% were females, the mean age was 44.5 years, and 45% majored in social work as undergraduates. Individuals who reportedly had ‘a great deal of influence’ in their selection of graduate social work education included social work practitioners, family members, and undergraduate professors. The ‘very important’ EDFs included geographical location, type of social work program, class scheduling, and cost of tuition. The strongest EDF correlations were between (1) number of minority students and number of minority faculty (r?=?0.662, p?=?0.000, 0.01 level, 2‐tailed); (2) academic reputation and type of social work program (r?=?0.590, p?=?0.000, 0.01 level, 2‐tailed); and (3) number of minority faculty and school climate toward minorities (r?=?0.523, p?=?0.000, 0.01 level, 2‐tailed). Findings suggest that graduate school recruitment strategies for African Americans and other under‐represented groups should include use of social work practitioners and educators, as well as more informed and deliberate marketing approaches with respect to school curricula and how it translates into future career options.     

Are People Emotionally Predisposed to Experience Lower Quality of Life? The Impact of Negative Affectivity on Quality of Life in Patients Recovering from Cardiac Surgery

Negative affectivity has been defined as a predisposition to experience intense states of negative emotions. As a trait concept it is a dimension that reflects stable and pervasive differences in negative mood and self-concept. There has been systematic evidence linking negative affectivity to anxiety, depression, psychosomatic complaints, pain symptoms, and adherence to life style changes. Several studies have demonstrated a link between negative affectivity and quality of life. However, as the majority of these studies have been cross-sectional, it has been difficult to establish causal relations between quality of life and negative affect. The present study is a longitudinal investigation of negative affectivity and quality of life in patients undergoing coronary artery bypass grafting (N = 157). Negative affectivity and quality of life were assessed with the Positive and Negative Affect Schedule (PANAS) and with the MacNew Quality of Life after Myocardial Infarction Questionnaire (MacNew QLMI) on admission day, one, and six months after the operation. Causal and reciprocal relationships were tested using a cross-lagged designs analysis via structural equation modeling. Results indicated that the causal model represented the best fit to the data. Results of the study are discussed in terms of the clinical importance of identifying patients at risk of experiencing severe restrictions in their quality of life.     

Seasonal Patterns of Reproduction in Matlab,Bangladesh

In Bangladesh twice as many births occur in December as in July. This paper examines the seasonal patterns of the risks of conception, fetal loss and return of menses post partum in a longitudinal study of 2,300 women in 14 villages of Matlab, Bangladesh. Life tables were estimated for each month of entry event and then ‘period’ life tables were constructed with the risks for a given calendar month. Confirming the results of earlier studies, risks of resumption of menses were higher in November and December, regardless of time elapsed since the last birth. Similarly, there are increased conception risks in the period from February to April for all fecundable women. The time of lowest fetal loss and stillbirth risks is in the cool season, though this variation makes only a minor contribution to the overall seasonality of births. The pattern of fecundability estimated from data on coital frequency did not match the pattern estimated from reported conceptions; these discrepancies imply possible seasonal changes in other parameters of fecundability besides intercourse.