Intraday volatility interaction between the crude oil and equity markets

This paper investigates the price volatility interaction between the crude oil and equity markets in the US using 5-min data over the period 2009-2012. Our main findings can be summarised as follows. First, we find strong evidence to demonstrate that the integration of the bid-ask spread and trading volume factors leads to a better performance in predicting price volatility. Second, trading information, such as bid-ask spread, trading volume, and the price volatility from cross-markets, improves the price volatility predictability for both in-sample and out-of-sample analyses. Third, the trading strategy based on the predictive regression model that includes trading information from both markets provides significant utility gains to mean-variance investors.

Determinants of mastitis in women in the CASTLE study: a cohort study

BACKGROUND: Mastitis is an acute, debilitating condition that occurs in approximately 20 % of breastfeeding women who experience a red, painful breast with fever. This paper describes the factors correlated with mastitis and investigates the presence of Staphylococcus aureus in women who participated in the CASTLE (Candida and Staphylococcus Transmission: Longitudinal Evaluation) study. The CASTLE study was a prospective cohort study which recruited nulliparous women in late pregnancy in two maternity hospitals in Melbourne, Australia in 2009-2011.

METHODS: Women completed questionnaires at recruitment and six time-points in the first eight weeks postpartum. Postpartum questionnaires asked about incidences of mastitis, nipple damage, milk supply, expressing practices and breastfeeding problems. Nasal and nipple swabs were collected from mothers and babies, as well as breast milk samples. All samples were cultured for S. aureus. "Time at risk" of mastitis was defined as days between birth and first occurrence of mastitis (for women who developed mastitis) and days between birth and the last study time-point (for women who did not develop mastitis). Risk factors for incidence of mastitis occurring during the time at risk (Incident Rate Ratios [IRR]) were investigated using a discrete version of the multivariable proportional hazards regression model.

RESULTS: Twenty percent (70/346) of participants developed mastitis. Women had an increased risk of developing mastitis if they reported nipple damage (IRR 2.17, 95 % CI 1.21, 3.91), over-supply of breast milk (IRR 2.60, 95 % CI 1.58, 4.29), nipple shield use (IRR 2.93, 95 % CI 1.72, 5.01) or expressing several times a day (IRR 1.64, 95 % CI 1.01, 2.68). The presence of S. aureus on the nipple (IRR 1.72, 95 % CI 1.04, 2.85) or in milk (IRR 1.78, 95 % CI 1.08, 2.92) also increased the risk of developing mastitis.

CONCLUSIONS: Nipple damage, over-supply of breast milk, use of nipple shields and the presence of S. aureus on the nipple or in breast milk increased the mastitis risk in our prospective cohort study sample. Reducing nipple damage may help reduce maternal breast infections.

Maternal depressive and anxiety symptoms, self-esteem, body dissatisfaction and preschooler obesity: a cross-sectional study

The primary aim of the present study was to cross-sectionally examine the associations between maternal psychosocial variables, child feeding practices, and preschooler body mass index z-score (BMI-z) in children (aged 2–4 years). A secondary aim was to examine differences in child weight outcomes between mothers scoring above and below specified cut-offs on the psychosocial measures. Two hundred and ninety mother–child dyads were recruited from Melbourne, Australia, and completed questionnaires examining demographic information, mothers’ depressive and anxiety symptoms, self-esteem and body dissatisfaction, restrictive and pressure child feeding practices, and preschoolers’ BMI-z scores. Independent t-tests and hierarchical multiple regression were employed to analyse the data. In the final regression model, none of the maternal psychosocial measures or feeding practices predicted child BMI-z scores; maternal body mass index and employment status were the only predictors of preschooler BMI-z. However, independent t-tests revealed that children of mothers with elevated body dissatisfaction scores had significantly higher BMI-z scores than children of mothers without elevated scores. The results suggest that psychosocial variables are not related, cross-sectionally, to preschooler weight outcomes; however, further research is needed to replicate the group differences noted between mothers with and without body dissatisfaction, and to track these relationships longitudinally.

Anxiety, depression, impulsivity and substance misuse in violent and non-violent adolescent boys in detention in China

The present investigation aims to identify the factors which differentiate violent from non-violent juvenile offenders, with a particular emphasis on the association between internalizing psychiatric morbidity (i.e. anxiety and depression), impulsivity, substance misuse, and violence. A total of 323 incarcerated male juvenile offenders from one of three Youth Detention Centers (YDCs) in China were recruited between August 2007 and November 2008. Interviews were conducted by trained psychiatrists using the Barratt Impulsivity Scale (BIS-11), the Screen for Child Anxiety Related Emotional Disorders (SCARED), and the Birleson Depression Self-Rating Scale (DSRS) to assess impulsivity, anxiety and depression, respectively. The Schedule for Affective Disorder and Schizophrenia for School-Age Children Present and Lifetime (K-SADS-PL) was also used to assess psychiatric diagnoses. Violent offenders had significantly higher BIS-11 total scores, and attention and nonplanning subscale scores (p<0.05). In the multiple logistic regression model, substance use disorders (SUD) and BIS-11 total scores independently predicted violence. Prison-based treatment services designed to reduce impulsivity and substance misuse in juvenile detention facilities should be prioritized.

Stabilizing high-dimensional prediction models using feature graphs

We investigate feature stability in the context of clinical prognosis derived from high-dimensional electronic medical records. To reduce variance in the selected features that are predictive, we introduce Laplacian-based regularization into a regression model. The Laplacian is derived on a feature graph that captures both the temporal and hierarchic relations between hospital events, diseases, and interventions. Using a cohort of patients with heart failure, we demonstrate better feature stability and goodness-of-fit through feature graph stabilization.

Approximation of centroid end-points and switch points for replacing type reduction algorithms

Despite several years of research, type reduction (TR) operation in interval type-2 fuzzy logic system (IT2FLS) cannot perform as fast as a type-1 defuzzifier. In particular, widely used Karnik-Mendel (KM) TR algorithm is computationally much more demanding than alternative TR approaches. In this work, a data driven framework is proposed to quickly, yet accurately, estimate the output of the KM TR algorithm using simple regression models. Comprehensive simulation performed in this study shows that the centroid end-points of KM algorithm can be approximated with a mean absolute percentage error as low as 0.4%. Also, switch point prediction accuracy can be as high as 100%. In conjunction with the fact that simple regression model can be trained with data generated using exhaustive defuzzification method, this work shows the potential of proposed method to provide highly accurate, yet extremely fast, TR approximation method. Speed of the proposed method should theoretically outperform all available TR methods while keeping the uncertainty information intact in the process.

Visual impairment and socioeconomic factors

BACKGROUND: Information about socioeconomic factors associated with visual impairment can assist in the design of intervention programmes. Such information was collected by the Melbourne Visual Impairment Project (Melbourne VIP). METHODS: The Melbourne VIP was a population based study of non-institutionalised permanent residents in nine suburbs of the Melbourne metropolitan area aged 40 years of age and older. A standardised eye examination was provided to eligible residents which included a structured interview. Variables of interest for this analysis were age, sex, country of birth, language spoken at home, education level, use of private health insurance, employment status, and living arrangements. Visual impairment was defined as a best corrected visual acuity < 6/18 and/or visual field constriction to within 20 degrees of fixation. RESULTS: A total of 3271 (83%) residents participated. Participants ranged in age from 40 to 98 years; 54% were female. Forty four (1.34%) were classified as visually impaired due to visual acuity and/or visual field loss. To evaluate the independent association of the significant sociodemographic variables with visual impairment, a regression model was constructed that included age, retirement status, use of private health insurance, and household arrangement. The results showed that age was the significant predictor of visual impairment (OR: 3.19; CI: 2.29-4.43), with the mean age of people with visual impairment significantly older (75.0 years) compared with people without visual impairment (58.2 years) (t test = 9.71; p = 0.0001). Of the 44 visually impaired people, 39 (87%) were aged 60 years of age and older. CONCLUSION: The results indicate that age is the most significant factor associated with visual impairment. Of some importance was the finding that people with visual impairment were less likely to have private health insurance. With the aging of the population, the number of people affected by visual impairment will increase significantly. Intervention programmes need to be established before the onset of middle age to offset the escalation of visual impairment in the older population.

Fusing data mining, machine learning and traditional statistics to detect biomarkers associated with depression

BACKGROUND: Atheoretical large-scale data mining techniques using machine learning algorithms have promise in the analysis of large epidemiological datasets. This study illustrates the use of a hybrid methodology for variable selection that took account of missing data and complex survey design to identify key biomarkers associated with depression from a large epidemiological study.

METHODS: The study used a three-step methodology amalgamating multiple imputation, a machine learning boosted regression algorithm and logistic regression, to identify key biomarkers associated with depression in the National Health and Nutrition Examination Study (2009-2010). Depression was measured using the Patient Health Questionnaire-9 and 67 biomarkers were analysed. Covariates in this study included gender, age, race, smoking, food security, Poverty Income Ratio, Body Mass Index, physical activity, alcohol use, medical conditions and medications. The final imputed weighted multiple logistic regression model included possible confounders and moderators.

RESULTS: After the creation of 20 imputation data sets from multiple chained regression sequences, machine learning boosted regression initially identified 21 biomarkers associated with depression. Using traditional logistic regression methods, including controlling for possible confounders and moderators, a final set of three biomarkers were selected. The final three biomarkers from the novel hybrid variable selection methodology were red cell distribution width (OR 1.15; 95% CI 1.01, 1.30), serum glucose (OR 1.01; 95% CI 1.00, 1.01) and total bilirubin (OR 0.12; 95% CI 0.05, 0.28). Significant interactions were found between total bilirubin with Mexican American/Hispanic group (p = 0.016), and current smokers (p<0.001).

CONCLUSION: The systematic use of a hybrid methodology for variable selection, fusing data mining techniques using a machine learning algorithm with traditional statistical modelling, accounted for missing data and complex survey sampling methodology and was demonstrated to be a useful tool for detecting three biomarkers associated with depression for future hypothesis generation: red cell distribution width, serum glucose and total bilirubin.