DeepSoft: a vision for a deep model of software

Although software analytics has experienced rapid growth as a research area, it has not yet reached its full potential for wide industrial adoption. Most of the existing work in software analytics still relies heavily on costly manual feature engineering processes, and they mainly address the traditional classification problems, as opposed to predicting future events. We present a vision for \emph{DeepSoft}, an \emph{end-to-end} generic framework for modeling software and its development process to predict future risks and recommend interventions. DeepSoft, partly inspired by human memory, is built upon the powerful deep learning-based Long Short Term Memory architecture that is capable of learning long-term temporal dependencies that occur in software evolution. Such deep learned patterns of software can be used to address a range of challenging problems such as code and task recommendation and prediction. DeepSoft provides a new approach for research into modeling of source code, risk prediction and mitigation, developer modeling, and automatically generating code patches from bug reports.

The conceptualisation of operational risk management models

Operational risk is evolving as a specialist field of risk management that must be practiced within all organisations, but currently has a particular relevance to banks. The Basel Committee on Banking Supervision has circulated a consultative paper which, if adopted by nation-state bank supervisors, will impose an operational risk capital charge on banks as part of a new Capital Accord. The definition of operational risk is wide-ranging and creates some unique issues related to the development of appropriate risk management models. This paper conceptualises two distinct operational risk management models; being a predictive model that will result in a known outcome upon its implementation, and a pre-emptive operational risk management model which prepares an organisation in the event that a future risk occurrence results in a disruption to critical business operations.

Suicide among psychiatric patients : a case-control study

Objective: To examine patient- and treatment-based differences between psychiatric patients who do and do not die by suicide. Method: By linking databases of deaths and psychiatric service use in Victoria, we compared 597 cases who suicided over 5 years with individually matched controls. Results: Cases and controls could not be distinguished on the majority of patient- or treatment- based characteristics. The exceptions were that cases were more likely to be male, less likely to be outside the labour force, more likely to have recent contact with inpatient and community services, and more likely to have a registration as their last contact. Conclusions: Patients who suicide 'look' similar to those who do not, suggesting prevention approaches should ensure that all psychiatric patients receive optimal care, including appropriate detection, diagnosis, assessment and treatment of mental health problems, and careful, individualised assessment of suicide risk.

The conceptualisation of operational risk management models

Operational risk is evolving as a specialist field of risk management that must be practiced within all organisations, but currently has a particular relevance to banks. The Basel Committee on Banking Supervision has circulated a consultative paper which, if adopted by nation-state bank supervisors, will impose an operational risk capital charge on banks as part of a new Capital Accord. The definition of operational risk is wide-ranging and creates some unique issues related to the development of appropriate risk management models. This paper conceptualises two distinct operational risk management models; being a predictive model that will result in a known outcome upon its implementation, and a pre-emptive operational risk management model which prepares an organisation in the event that a future risk occurrence results in a disruption to critical business operations.

The metabolic syndrome as a tool for predicting future diabetes : the AusDiab study

Objective. To compare the ability of the metabolic syndrome (MetS), a diabetes prediction model (DPM), a noninvasive risk questionnaire and individual glucose measurements to predict future diabetes.

Design. Five-year longitudinal cohort study. Tools tested included MetS definitions [World Health Organization, International Diabetes Federation, ATPIII and European Group for the study of Insulin Resistance (EGIR)], the FINnish Diabetes RIsk SCore risk questionnaire, the DPM, fasting and 2-h post load plasma glucose.

Setting. Adult Australian population.

Subjects. A total of 5842 men and women without diabetes ≥25 years. Response 58%. A total of 224 incident cases of diabetes.

Results. In receiver operating characteristic curve analysis, the MetS was not a better predictor of incident diabetes than the DPM or measurement of glucose. The risk for diabetes among those with prediabetes but not MetS was almost triple that of those with MetS but not prediabetes (9.0% vs. 3.4%). Adjusted for component parts, the MetS was not a significant predictor of incident diabetes, except for EGIR in men [OR 2.1 (95% CI 1.2–3.7)].

Conclusions. A single fasting glucose measurement may be more effective and efficient than published definitions of the MetS or other risk constructs in predicting incident diabetes. Diagnosis of the MetS did not confer increased risk for incident diabetes independent of its individual components, with an exception for EGIR in men. Given these results, debate surrounding the public health utility of a MetS diagnosis, at least for identification of incident diabetes, is required.

A social network analysis model for analysing stakeholder-associated safety risks in infrastructure projects

An important strategy in the long-term blueprint for making Australia's 18 capital and major regional cities more productive, sustainable and liveable is to develop high quality public infrastructure systems to improve civic quality of life. Because of the unique features of construction activities, such as long period, complicated processes, and dynamic organizational structures, infrastructure projects normally involve multiple stakeholders and are subject to various risks, especially safety issues. Any negligence or mismanagement of critical safety risks will have huge impact on achieving project objectives and success. Although many previous studies have identified and assessed various safety risks in construction industry, a main research gap is that these studies ignored a fact that most risks are interrelated and associated with internal and external stakeholders of the projects. The lack of a theoretical foundation and appropriate methods for analysing stakeholder-associated safety risks and their interdependencies in infrastructure projects hinders effective risk management processes and the formulations of decision strategies. This research aims at enabling higher performance in strategic safety risk management in infrastructure projects through the development of a holistic risk analysis model using Stakeholder and Social Network Theories. The outcomes can broaden project managers' awareness of emerging influential safety risks and enhance their ability to perceive, understand, assess, and mitigate safety risks in an effective and efficient way; thereby higher performance in strategic risk management could be achieved in infrastructure projects.

The influence of hip circumference on the relationship between abdominal obesity and mortality

Background Higher waist circumference and lower hip circumference are both associated with increased cardiovascular disease (CVD) risk, despite being directly correlated. The real effects of visceral obesity may therefore be underestimated when hip circumference is not fully taken into account. We hypothesized that adding waist and hip circumference to traditional risk factors would significantly improve CVD risk prediction.

Methods In a population-based survey among South Asian and African Mauritians (n = 7978), 1241 deaths occurred during 15 years of follow-up. In a model that included variables used in previous CVD risk calculations (a Framingham-type model), the association between waist circumference and mortality was examined before and after adjustment for hip circumference. The percentage with an increase in estimated 10-year cumulative mortality of >25% and a decrease of >20% after waist and hip circumference were added to the model was calculated.

Results Waist circumference was strongly related to mortality only after adjustment for hip circumference and vice versa. Adding waist and hip circumference to a Framingham-type model increased estimated 10-year cumulative CVD mortality by >25% for 23.7% of those who died and 15.7% of those censored. Cumulative mortality decreased by >20% for 4.5% of those who died and 14.8% of those censored.

Conclusions The effect of central obesity on mortality risk is seriously underestimated without adjustment for hip circumference. Adding waist and hip circumference to a Framingham-type model for CVD mortality substantially increased predictive power. Both may be important inclusions in CVD risk prediction models.

SAS macros for point and interval estimation of area under the receiver operating characteristic curve for non-proportional and proportional hazards Weibull models

Aims and objectives  For prediction of risk of cardiovascular end points using survival models the proportional hazards assumption is often not met. Thus, non-proportional hazards models are more appropriate for developing risk prediction equations in such situations. However, computer program for evaluating the prediction performance of such models has been rarely addressed. We therefore developed SAS macro programs for evaluating the discriminative ability of a non-proportional hazards Weibull model developed by Anderson (1991) and that of a proportional hazards Weibull model using the area under receiver operating characteristic (ROC) curve.

Method  Two SAS macro programs for non-proportional hazards Weibull model using Proc NLIN and Proc NLP respectively and model validation using area under ROC curve (with its confidence limits) were written with SAS IML language. A similar SAS macro for proportional hazards Weibull model was also written.

Results  The computer program was applied to data on coronary heart disease incidence for a Framingham population cohort. The five risk factors considered were current smoking, age, blood pressure, cholesterol and obesity. The predictive ability of the non-proportional hazard Weibull model was slightly higher than that of its proportional hazard counterpart. An advantage of SAS Proc NLP in terms of the example provided here is that it provides significance level for the parameter estimates whereas Proc NLIN does not.

Conclusion  The program is very useful for evaluating the predictive performance of non-proportional and proportional hazards Weibull models.

Assessing risk post intervention for an acute coronary syndrome: a review of the risk assessment tools and their development

Considerable variability in survival rate after an acute myocardial infarction exists and accurate risk stratification is of significant importance. The American College of Cardiology and the American Heart Association has recommended early risk stratification using several clinical risk scoring instruments to identify high risk patients. The aim of this paper is to identify secondary cardiovascular risk scoring instruments that could be utilized at the time of intervention for acute coronary syndromes and compare their psychometric properties as they were developed. A search using Medline, Cumulative Index to Nursing and Allied Health Literature and the Psychology and Behavioral Sciences Collection data-bases identified studies published between January 1990 and January 2010 used to measure risk after intervention for acute coronary syndrome. Four validated secondary risk prediction scoring instruments were identified for comparison.Secondary risk prediction scoring instruments for the acute coronary syndrome patient population are evidence based, valid and reliable. Use of the instruments by cardiac focused clinicians will aid in the determination of treatment strategies, and estimation of short and long term events and mortality.

Safety and acceptability of practice-nurse-managed care of depression in patients with diabetes or heart disease in the Australian TrueBlue study

Objectives:
To determine the safety and acceptability of the TrueBlue model of nurse-managed care in the primary healthcare setting.

Patient admission prediction using a pruned fuzzy min–max neural network with rule extraction

A useful patient admission prediction model that helps the emergency department of a hospital admit patients efficiently is of great importance. It not only improves the care quality provided by the emergency department but also reduces waiting time of patients. This paper proposes an automatic prediction method for patient admission based on a fuzzy min–max neural network (FMM) with rules extraction. The FMM neural network forms a set of hyperboxes by learning through data samples, and the learned knowledge is used for prediction. In addition to providing predictions, decision rules are extracted from the FMM hyperboxes to provide an explanation for each prediction. In order to simplify the structure of FMM and the decision rules, an optimization method that simultaneously maximizes prediction accuracy and minimizes the number of FMM hyperboxes is proposed. Specifically, a genetic algorithm is formulated to find the optimal configuration of the decision rules. The experimental results using a large data set consisting of 450740 real patient records reveal that the proposed method achieves comparable or even better prediction accuracy than state-of-the-art classifiers with the additional ability to extract a set of explanatory rules to justify its predictions.

Does gender explain the relationship between occupation and suicide? Findings from a meta-analytic study

This report investigated whether suicide risk by occupational groups differed for males and females. We examined this using a sub-set of articles examined in a recent meta-analysis and stratified by gender. For certain occupational groups, males and females had a similar risk of suicide (the military, community service occupations, managers, and clerical workers). There was some indication of gender differences for other occupations (technicians, plant and machine operators and ship’s deck crew, craft and related trades workers, and professionals), although these did not reach statistical significance. These findings highlight the complexity of the relationship between occupation and suicide and suggest the possible role of a range of individual, work-related and social-environmental risk factors that may differ for males and females.

Learning vector representation of medical objects via EMR-driven nonnegative restricted Boltzmann machines (eNRBM)

Electronic medical record (EMR) offers promises for novel analytics. However, manual feature engineering from EMR is labor intensive because EMR is complex - it contains temporal, mixed-type and multimodal data packed in irregular episodes. We present a computational framework to harness EMR with minimal human supervision via restricted Boltzmann machine (RBM). The framework derives a new representation of medical objects by embedding them in a low-dimensional vector space. This new representation facilitates algebraic and statistical manipulations such as projection onto 2D plane (thereby offering intuitive visualization), object grouping (hence enabling automated phenotyping), and risk stratification. To enhance model interpretability, we introduced two constraints into model parameters: (a) nonnegative coefficients, and (b) structural smoothness. These result in a novel model called eNRBM (EMR-driven nonnegative RBM). We demonstrate the capability of the eNRBM on a cohort of 7578 mental health patients under suicide risk assessment. The derived representation not only shows clinically meaningful feature grouping but also facilitates short-term risk stratification. The F-scores, 0.21 for moderate-risk and 0.36 for high-risk, are significantly higher than those obtained by clinicians and competitive with the results obtained by support vector machines.