Bayesian object matching

A Bayesian approach to object matching is presented. An object and a scene are each represented by features, such as critical points, line segments and surface patches, constrained by unary properties and contextual relations. The matching is presented as a labeling problem, where each feature in the scene is assigned (associated with) a feature of the known model objects. The prior distribution of a scene's labeling is modeled as a Markov random field, which encodes the between-object constraints. The conditional distribution of the observed features labeled is assumed to be Gaussian, which encodes the within-object constraints. An optimal solution is defined as a maximum a posteriori estimate. Relationships with previous work are discussed. Experimental results are shown.     

Transcending the fissure in risk theory: critical realism and child welfare

Within risk theory a key fissure has emerged. On the one hand, objectivist theories continue to treat risk as a measurable entity, while on the other, subjectivist accounts approach risk as a socially constructed phenomenon. Arguably, we need to transcend this false opposition and harmonize important insights from both approaches. Building on an earlier paper in this journal, the author tries to achieve this synthesis by applying Bhaskar's critical realist perspective to this contested area. The reconciliation of objectivism and subjectivism within critical realism, it is argued, redirects our attention to the deep‐seated causes of harm or the underlying mechanisms that, when activated, give rise to situations involving risk. The paper concludes by reviewing the implications of these ideas for child and family social work. It is here that connections are made with attachment theory and the recent interest in childhood resilience.     

ANALYZING NARRATIVES OF EXPERTISE:

Recent work in the sociology of professions highlights the central importance of abstract discourse in professionalization processes. Drawing on the work of Kenneth Burke, I argue that broadening the focus of analysis from "abstract discourse" to "narratives of expertise" will provide the ability to (1) more clearly analyze the social conditions that are conducive to the efficacy of abstraction as a basis for a claim of expertise and (2) theorize and empirically examine the formation and maintenance of a collective professional identity. To assist in the reformulation of professionalization studies around narratives of expertise, I develop a modified extension of Burke's five key terms of dramatism (act, agent, scene, agency, and purpose). Burke's pentadic scheme enables researchers to show how narratives of expertise ground jurisdictional claims and the constitution of professional identity in one of the elements of pentad. I illustrate the empirical power of the pertadic scheme through an examination of claims of psychotherapeutic expertise, particularly claims of marital and familial expertise.     

Who supports physicians in malpractice cases?

Malpractice suits have a devastating emotional impact on physicians. Learn about a unique service offered by the HealthTexas Provider Network to help doctors deal with the shock and sadness.     

An Improved Condensation Procedure in Discrete Probability Distribution Calculations

A "vertical" condensation scheme for discrete probability distribution (DPD) calculations is presented as an alternative to the earlier "horizontal" scheme, an example of which was presented recently by Kurth and Cox. When applied to DPDs over a space of curves, the vertical condensation results in a "regularization" of the "spaghetti" of curves that results from combination operations on such DPDs.     

Policy Development for the Reduction of Alcohol and Other Drug Problems at an Australian University

Universities, as with other workplaces, experience problems related to alcohol and other drug use by staff and students. There is little published information about the prevalence of these problems among students in the university setting in Australia as compared with the availability of information from the USA. Curtin University of Technology's experience with alcohol-related problems seems to be documented better than most other Australian institutions, and the available evidence indicates that the level of student alcohol consumption and reported alcohol-related problems is cause for concern. However, this is not to say that Curtin's problems may be any greater than those of any other Australian university with alcohol available on the premises. Many years of involvement in health promotion teaching and research, combined with identification of alcohol and drug related problems on campus, were thought to be the primary factors preceding the university's decision to begin the process of formulating comprehensive alcohol and other drug policy for staff and students. The absence of clear and consistent alcohol and other drug policy can lead to 1) undermining of the educational environment, 2) confusion about appropriate behaviour, 3) dealing with drug-related problems by inaction or arbitrary and inappropriate action. Alcohol and other drug policy has two main goals: problem reduction and better management of problems.     

Predictor relative importance and matching regression parameters

Predictor importance in applied regression modeling gives the main operational tools for managers and decision-makers. The paper considers estimation of predictors' importance in regression using measures introduced in works by Gibson and R. Johnson (GJ), then modified by Green, Carroll, and DeSarbo, and developed further by J. Johnson (JJ). These indices of importance are based on the orthonormal decomposition of the data matrix, and the work shows how to improve this approximation. Using predictor importance, the regression coefficients can also be adjusted to reach the best data fit and to be meaningful and interpretable. The results are compared with the robust to multicollinearity, but computationally difficult, Shapley value regression (SVR). They show that the JJ index is good for importance estimation, but the GJ index outperforms it if both predictor importance and coefficients of regression are needed; hence, this index (GJ) can be used in place of the more computationally intensive estimation by SVR. The results can be easily estimated by the considered approach that is very useful in practical regression modeling and analysis, especially for big data.     

Birth Intervals and Childhood Mortality in Rural Bangladesh

This study investigates the relationship between birth intervals and childhood mortality, using longitudinal data from rural Bangladesh known to be of exceptional accuracy and completeness. Results demonstrate significant but very distinctive effects of the previous and subsequent birth intervals on mortality, with the former concentrated in the neonatal period and the latter during early childhood. The impact of short birth intervals on mortality, however, is substantially less than that found in many previous studies of this issue, particularly for the previous birth interval. The findings are discussed in terms of the potential for family planning programs to contribute to improved child survival in settings such as Bangladesh.