Regression Techniques (regression + techniques)

Distribution by Scientific Domains
Distribution within Medical Sciences

Kinds of Regression Techniques

  • linear regression techniques
  • logistic regression techniques


  • Selected Abstracts


    More evidence of the need for an ergonomic standard

    AMERICAN JOURNAL OF INDUSTRIAL MEDICINE, Issue 4 2004
    Jeff Biddle
    Abstract Background In 1999, the Occupational Safety and Health Administration (OSHA) proposed regulations designed to reduce work related inquiries by limiting worker exposure to "ergonomic risk factors." Congress subsequently overturned the regulations. We provide additional evidence on earnings losses attributable to musculoskeletal disorders (MSDs), and thus on the need for an ergonomic standard. Methods Regression techniques are used to analyze data from a survey of injured workers that has been matched to employer-reported earnings data covering pre- and post-injury periods, and to workers' compensation claims records. Results MSDs lead to large and persistent earnings losses. Cost estimates used by OSHA to justify the 1999 EPS are corroborated. Losses are greatest among workers who file workers compensation claims, but nonclaimants also have losses. Conclusions Earnings losses and lost productivity associated with work-related MSDs are substantial and an ergonomic standard could be cost effective. Am. J. Ind. Med. 45:329,337, 2004. © 2004 Wiley-Liss, Inc. [source]


    The causes, consequences and detection of publication bias in psychiatry

    ACTA PSYCHIATRICA SCANDINAVICA, Issue 4 2000
    Simon M. Gilbody
    Objective: Publication bias threatens the validity of published research, although this topic has received little attention in psychiatry. The purpose of this article is to produce a systematic overview of the causes and consequences of publication bias and to summarize the available methods with which it is detected and corrected. Method: Empirical evidence for the existence of publication bias is reviewed and the following methods are applied to an illustrative case example from psychiatry: funnel plot analysis; the ,file drawer method'; linear regression techniques; rank correlation; ,trim and fill'. Results: Small studies are particularly susceptible to publication and related bias. All methods to detect publication bias depend upon the availability of a number of individual studies with a range of sample sizes. Unfortunately, large numbers of studies of varying sample size are not always available in many areas of psychiatric research. Conclusion: Where possible researchers should always test for the presence of publication bias. The problem of publication bias will not be solved by anything other than a prospective trials register. [source]


    Predicting species distributions from museum and herbarium records using multiresponse models fitted with multivariate adaptive regression splines

    DIVERSITY AND DISTRIBUTIONS, Issue 3 2007
    Jane Elith
    ABSTRACT Current circumstances , that the majority of species distribution records exist as presence-only data (e.g. from museums and herbaria), and that there is an established need for predictions of species distributions , mean that scientists and conservation managers seek to develop robust methods for using these data. Such methods must, in particular, accommodate the difficulties caused by lack of reliable information about sites where species are absent. Here we test two approaches for overcoming these difficulties, analysing a range of data sets using the technique of multivariate adaptive regression splines (MARS). MARS is closely related to regression techniques such as generalized additive models (GAMs) that are commonly and successfully used in modelling species distributions, but has particular advantages in its analytical speed and the ease of transfer of analysis results to other computational environments such as a Geographic Information System. MARS also has the advantage that it can model multiple responses, meaning that it can combine information from a set of species to determine the dominant environmental drivers of variation in species composition. We use data from 226 species from six regions of the world, and demonstrate the use of MARS for distribution modelling using presence-only data. We test whether (1) the type of data used to represent absence or background and (2) the signal from multiple species affect predictive performance, by evaluating predictions at completely independent sites where genuine presence,absence data were recorded. Models developed with absences inferred from the total set of presence-only sites for a biological group, and using simultaneous analysis of multiple species to inform the choice of predictor variables, performed better than models in which species were analysed singly, or in which pseudo-absences were drawn randomly from the study area. The methods are fast, relatively simple to understand, and useful for situations where data are limited. A tutorial is included. [source]


    Evaluation of coalbed methane reservoirs from geophysical log data using an improved fuzzy comprehensive decision method and a homologous neural network

    GEOPHYSICAL PROSPECTING, Issue 5 2002
    J. Hou
    The evaluation of coalbed methane reservoirs using log data is an important approach in the exploration and development of coalbed methane reservoirs. Most commonly, regression techniques, fuzzy recognition and neural networks have been used to evaluate coalbed methane reservoirs. It is known that a coalbed is an unusual reservoir. There are many difficulties in using regression methods and empirical qualitative recognition to evaluate a coalbed, but fuzzy recognition, such as the fuzzy comprehensive decision method, and neural networks, such as the back-propagation (BP) network, are widely used. However, there are no effective methods for computing weights for the fuzzy comprehensive decision method, and the BP algorithm is a local optimization algorithm, easily trapped in local minima, which significantly affect the results. In this paper, the recognition method for coal formations is developed; the improved fuzzy comprehensive decision method, which uses an optimization approach for computing weighted coefficients, is developed for the qualitative recognition of coalbed methane reservoirs. The homologous neural network, using a homologous learning algorithm, which is a global search optimization, is presented for the quantitative analysis of parameters for coalbed methane reservoirs. The applied procedures for these methods and some problems related to their application are also discussed. Verification of the above methods is made using log data from the coalbed methane testing area in North China. The effectiveness of the methods is demonstrated by the analysis of results for real log data. [source]


    Does Satisfaction Reflect the Technical Quality of Mental Health Care?

    HEALTH SERVICES RESEARCH, Issue 2 2003
    Mark J. Edlund
    Objective. To analyze the relationship between satisfaction and technical quality of care for common mental disorders. Data Source. A nationally representative telephone survey of 9,585 individuals conducted in 1997,1998. Study Design. Using multinomial logistic regression techniques we investigated the association between a five-level measure of satisfaction with the mental health care available for personal or emotional problems and two quality indicators. The first measure, appropriate technical quality, was defined as use of either appropriate counseling or psychotropic medications during the prior year for a probable depressive or anxiety disorder. The second, active treatment, indicated whether the respondent had received treatment for a psychiatric disorder in the past year. Covariates included measures of physical and mental health and sociodemographic indicators. Principal Findings. Appropriate technical quality of care was significantly associated with higher levels of satisfaction. The strength of the association was moderate. Conclusions. Satisfaction is associated with technical quality of care. However, profiling quality of care with satisfaction will likely require large samples and case-mix adjustment, which may be more difficult for plans or provider groups to implement than measuring technical indicators. More importantly, satisfaction is not the same as technical quality, and our results suggest that at this time they cannot be made to approach each other closely enough to eliminate either. [source]


    Brain cancer mortality and potential occupational exposure to lead: Findings from the National Longitudinal Mortality Study, 1979,1989

    INTERNATIONAL JOURNAL OF CANCER, Issue 5 2006
    Edwin van Wijngaarden
    Abstract We evaluated the association between potential occupational lead exposure and the risk of brain cancer mortality in the National Longitudinal Mortality Study (NLMS), which is a prospective census-based cohort study of mortality among the noninstitutionalized United States population (1979,1989). The present study was limited to individuals for whom occupation and industry were available (n = 317,968). Estimates of probability and intensity of lead exposure were assigned using a job-exposure matrix (JEM). Risk estimates for the impact of lead on brain cancer mortality were computed using standardized mortality ratio (SMR) and proportional hazards and Poisson regression techniques, adjusting for the effects of age, gender and several other covariates. Brain cancer mortality rates were greater among individuals in jobs potentially involving lead exposure as compared to those unexposed (age- and gender-adjusted hazard ratio (HR) = 1.5; 95% confidence interval (CI) = 0.9,2.3) with indications of an exposure,response trend (probability: low HR = 0.7 (95% CI = 0.2,2.2), medium HR = 1.4 (95% CI = 0.8,2.5), high HR = 2.2 (95% CI = 1.2,4.0); intensity: low HR = 1.2 (95% CI = 0.7,2.1), medium/high HR = 1.9 (95% CI = 1.0,3.4)). Brain cancer risk was greatest among individuals with the highest levels of probability and intensity (HR = 2.3; 95% CI = 1.3,4.2). These findings provide further support for an association between occupational lead exposure and brain cancer mortality, but need to be interpreted cautiously due to the consideration of brain cancer as one disease entity and the absence of biological measures of lead exposure. © 2006 Wiley-Liss, Inc. [source]


    Exploring consumer ethics in Ghana, West Africa

    INTERNATIONAL JOURNAL OF CONSUMER STUDIES, Issue 6 2007
    Samuel K. Bonsu
    Abstract Consumer ethics is a growing area of research that focused almost exclusively on consumers in the United States and, to a lesser degree, Europe and Asia. In this paper, we introduce an African element to the consumer ethics discourse by drawing on survey responses from over 300 Ghanaian consumers to explore their ethical beliefs and judgements. We analysed these data using regression techniques. Our findings show that Ghanaian consumers exhibit lower levels of ethics compared with their America counterparts, especially when the unethical actions facilitate the achievement of their goals. While Ghanaian consumers recognize the value of moral rules, they are prone to suspending their ethical positions as they deem necessary in a particular context. Implications for marketing strategy and future research are discussed. [source]


    Conoscopic holograms analysis using variations of the Hough transform

    INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY, Issue 1 2009
    Julius Butime
    Abstract The analysis of holograms obtained using a Conoscopic setup is a very complex subject, given their nature. Holograms contain the 3D depth information about an object whose surface is to be measured and later reconstructed. The recovery of depth information has so far been carried out using mathematical transforms in combination with linear regression techniques. Here the Hough transform, a useful Computer Vision technique for detecting features in images is adapted to the analysis of holograms in order to establish distance relationships for a given object. The captured images of holograms are pre-processed and subsequently analyzed for characteristic patterns that are later used in finding the distance to an object. © 2009 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 19, 10,13, 2009 [source]


    Penalized Regression with Ordinal Predictors

    INTERNATIONAL STATISTICAL REVIEW, Issue 3 2009
    Jan Gertheiss
    Summary Ordered categorial predictors are a common case in regression modelling. In contrast to the case of ordinal response variables, ordinal predictors have been largely neglected in the literature. In this paper, existing methods are reviewed and the use of penalized regression techniques is proposed. Based on dummy coding two types of penalization are explicitly developed; the first imposes a difference penalty, the second is a ridge type refitting procedure. Also a Bayesian motivation is provided. The concept is generalized to the case of non-normal outcomes within the framework of generalized linear models by applying penalized likelihood estimation. Simulation studies and real world data serve for illustration and to compare the approaches to methods often seen in practice, namely simple linear regression on the group labels and pure dummy coding. Especially the proposed difference penalty turns out to be highly competitive. Résumé Les variables indépendantes catégoriques ordinales sont un cas courant dans les modèles de régression. Contrairement au cas des variables dépendantes ordinales, les variables indépendantes ordinales ont été largement négligées par la recherche. Le présent article présente les méthodes existantes et propose l'utilisation de techniques de régression pénalisée. Deux types de pénalisation basés sur des variables dummy sont exposés; le premier impose une pénalité de différence, le second est une procédure basée sur une forme de régression ridge. D'autre part, une motivation baysienne est présentée. La méthode est également appliquée au cas de variables dépendantes non gaussiennes. Des études de simulation et des données réelles servent à illustrer et à comparer les nouvelles méthodes aux méthodes que l'on rencontre souvent dans la pratique - à savoir les régressions linéaires sur les nombres entiers et sur des variables dummy sans penalité. Une pénalité de différence notamment a montré de bons résultats. [source]


    ,To Whom Much Has Been Given...': Religious Capital and Community Voluntarism Among Churchgoing Protestants

    JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION, Issue 3 2000
    Jerry Z. Park
    Research on volunteering behavior has consistently found a positive relationship between religion and volunteering. Using a sample of churchgoing Protestants (N=1,738)from the Religious Identity and Influence Survey we examine the specific influences of religiosity, religious identity, religious socialization, and religious social networks on local volunteer activity in church programs and non-church organizations, as well as general volunteering tendencies. These influences are presented within the theoretical framework of religious capital. Logistic regression techniques were applied to determine the strength of the contribution of these influences while accounting for basic background factors. Findings suggest that churchgoing Protestants are influenced by all measures to some degree, but religiosity (specifically participation in church activities) remains the strongest influence. Significant religious influences overall are most pronounced within the context of church-related volunteering which suggests that churchgoing Protestants exhibit a strong sense of community identity through their local churches. A discussion of these results and their implications for volunteering follows. [source]


    The occurrence of commensal rodents in dwellings as revealed by the 1996 English House Condition Survey

    JOURNAL OF APPLIED ECOLOGY, Issue 4 2001
    S.D. Langton
    Summary 1,The presence of commensal rodents was assessed in the 1996 English House Condition Survey (EHCS). Logistic regression techniques were used to identify the key factors that might determine the susceptibility of dwellings to infestation. 2,The overall percentages of dwellings that were infested, weighted to allow for the more intensive sampling used in certain categories of dwellings, were 1·83% for mice Mus domesticus, 0·23% for rats Rattus norvegicus living indoors and 1·60% for rats living outdoors. These figures excluded vacant properties, properties with some commercial use, and purpose-built flats, as these groups showed different patterns of infestation and were therefore excluded from the logistic regressions. 3,The prevalence of both rats and mice was significantly greater for dwellings where pets or livestock were kept in the garden. 4,Dwellings classed as unfit for human habitation were more likely to be infested with mice. 5,Dwellings in areas of low-density housing had a significantly higher prevalence of both rat and mouse infestation. This probably reflects the general suitability of the rural environment for commensal rodents. 6,Older properties had a relatively high prevalence of rats. This may be because their mature gardens provided suitable habitats for colonization. Once other confounding factors were taken into account, the age of the property did not influence the rate of infestation by mice. 7,Dwellings in areas with substantial problems, such as dereliction, litter, vacant properties and unkempt gardens, had a significantly higher prevalence of rats and mice. 8,This study reveals the value of applied ecological techniques, including logistic regression of presence,absence data, in understanding the distribution of commensal rodents in relation to dwellings, with the prospect of more effective management practices being developed as a consequence. [source]


    Effects of ecogeographic variables on genetic variation in montane mammals: implications for conservation in a global warming scenario

    JOURNAL OF BIOGEOGRAPHY, Issue 7 2007
    Amy M. Ditto
    Abstract Aim, Evolutionary theory predicts that levels of genetic variation in island populations will be positively correlated with island area and negatively correlated with island isolation. These patterns have been empirically established for oceanic islands, but little is known about the determinants of variation on habitat islands. The goals of this study were twofold. Our first aim was to test whether published patterns of genetic variation in mammals occurring on montane habitat islands in the American Southwest conformed to expectations based on evolutionary theory. The second aim of this research was to develop simple heuristic models to predict changes in genetic variation that may occur in these populations as a result of reductions in available mountaintop habitat in response to global warming. Location, Habitat islands of conifer forest on mountaintops in the American Southwest. Methods, Relationships between island area and isolation with measures of allozyme variation in four species of small mammal, namely the least chipmunk (Tamias minimus), Colorado chipmunk (Tamias quadrivittatus), red squirrel (Tamiasciurus hudsonicus), and Mexican woodrat (Neotoma mexicana), were determined using correlation and regression techniques. Significant relationships between island area and genetic variation were used to develop three distinct statistical models with which to predict changes in genetic variation following reduction in insular habitat area arising from global warming. Results, Patterns of genetic variation in each species conformed to evolutionary predictions. In general, island area was the most important determinant of heterozygosity, while island isolation was the most important determinant of polymorphism and allelic diversity. The heuristic models predicted widespread reductions in genetic variation, the extent of which depended on the population and model considered. Main conclusions, The results support a generalized pattern of genetic variation for any species with an insular distribution, with reduced variation in smaller, more isolated populations. We predict widespread reductions in genetic variation in isolated populations of montane small mammals in the American Southwest as a result of global warming. We conclude that climate-induced reductions in the various dimensions of genetic variation may increase the probability of population extinction in both the short and long term. [source]


    Wear in molded tibial inserts: Knee simulator study of H1900 and GUR1050 polyethylenes

    JOURNAL OF BIOMEDICAL MATERIALS RESEARCH, Issue 2 2008
    Riichiro Tsukamoto
    Abstract Hi-fax 1900Ô tibial inserts were used in the IB-1 total knee replacement (TKR) beginning 1978, soon followed by the AGC design. Such direct compression molded (DCM) inserts was relatively immune to oxidation. Unfortunately the Hi-fax 1900Ô resin (H1900) was taken off the market in year 2004. As an alternate, GUR1050 was introduced in the VanguardÔ TKR. However there appeared to be little or no wear comparisons of molded inserts. Therefore the study aim was to compare wear performance of GUR1050 to the historical H1900. The hypothesis was that Hi-fax and GUR1050 would show comparable wear performance. The VanguardÔ was a posterior-cruciate sacrificing design (Biomet Inc.). All tibial inserts were sterilized by gamma-radiation (3.2 Mrad) under argon. A 6-channel, displacement,controlled knee simulator was used with serum lubricant (protein concentration 20 mg/mL). Wear assessments were by gravimetric methods and linear regression techniques. The gross weight-loss trends over 2.5 Mc duration demonstrated excellent linear behavior with good agreement between TKR sets (<±10%). Fluid sorption artifacts in control represented less than 5% of gross wear magnitudes. Thus suitable corrections could be made in determining net wear. The H1900 and GUR1050inserts demonstrated net wear-rates of 3.6 and 3.4 mm3/Mc, respectively. This difference was not found to be statistically significant. This wear study demonstrated that GUR1050 inserts were indistinguishable from the Hi-fax 1900 in terms of laboratory wear performance, proving our hypothesis. Given the excellent clinical history of DCM Hi-fax 1900, the GUR1050 should be an ideal candidate for TKR. © 2007 Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater, 2008 [source]


    Role of executive dysfunction in predicting frequency and severity of violence

    AGGRESSIVE BEHAVIOR, Issue 5 2010
    Megan Hancock
    Abstract The adverse consequences of violence on society are tremendous. The proportion of offenders incarcerated for violent offenses is large, and the cost of keeping these offenders incarcerated is startling. Understanding and treating the causal underpinnings of violent crime is of utmost importance for individuals and society as a whole. Several factors have been identified as potential contributors to violent crime, including cognitive deficits in executive functioning [Hoaken et al., 2007]. To investigate this further, 77 offenders from Fenbrook Institution, a federal facility, were tested on a battery of executive functioning measures. Offenders were found to have broad and pervasive dysfunction in their executive abilities. In addition, specific scores from the battery were found using regression techniques, to predict the frequency and severity of past violent offending but not nonviolent offending. This speaks of the possibility of a new type of correctional rehabilitation program, one that focuses on the rehabilitation of basic executive functions. Aggr. Behav. 36:338,349, 2010. © 2010 Wiley-Liss, Inc. [source]


    Spatial regression techniques for inter-population data: studying the relationships between morphological and environmental variation

    JOURNAL OF EVOLUTIONARY BIOLOGY, Issue 2 2010
    S. I. PEREZ
    Abstract Understanding the importance of environmental dimensions behind the morphological variation among populations has long been a central goal of evolutionary biology. The main objective of this study was to review the spatial regression techniques employed to test the association between morphological and environmental variables. In addition, we show empirically how spatial regression techniques can be used to test the association of cranial form variation among worldwide human populations with a set of ecological variables, taking into account the spatial autocorrelation in data. We suggest that spatial autocorrelation must be studied to explore the spatial structure underlying morphological variation and incorporated in regression models to provide more accurate statistical estimates of the relationships between morphological and ecological variables. Finally, we discuss the statistical properties of these techniques and the underlying reasons for using the spatial approach in population studies. [source]


    Farmers' health and agricultural productivity

    AGRICULTURAL ECONOMICS, Issue 4 2009
    Maria L. Loureiro
    Occupational health; Stochastic frontier regression; Norwegian agriculture Abstract Agriculture is one of the most hazardous productive sectors around the world. Most previous studies have focused on health issues of farmers in developing countries, while little attention has been paid to farmers' health problems in developed countries. The present study assesses the effect of farmers' health conditions on agricultural productivity in Norway. Employing stochastic frontier regression techniques, we conclude that differences in farmers' health help to explain the variance in agricultural production efficiency. [source]


    Gender, Work, and Intimate Violence: Men's Occupational Violence Spillover and Compensatory Violence

    JOURNAL OF MARRIAGE AND FAMILY, Issue 4 2002
    Scott A. Melzer
    Researchers have rarely studied the effects of occupations on intimate violence, only occasionally distinguishing between blue-collar and white-collar work, and generally finding higher rates of reported abuse in the former group. This research incorporates ideas from feminist, work-family, and power or resource theories to examine the potential effects of occupations on men's violence toward wives and cohabiting female partners. Data from the 1988 National Survey of Families and Households (NSFH) were analyzed using logistic regression techniques. Hypotheses related to occupational spillover and compensation were tested with results suggesting that men in physically violent, female-dominated, professional specialty, and dangerous occupations are more likely to use violence against female partners, net of other commonly hypothesized predictors. The findings suggest that more detailed occupational data should be collected in future intimate violence research. [source]


    Comparison of Adaptive Behavior in Children With Heavy Prenatal Alcohol Exposure or Attention-Deficit/Hyperactivity Disorder

    ALCOHOLISM, Issue 11 2009
    Nicole Crocker
    Background:, Adaptive behavior, the ability to respond successfully to everyday demands, may be especially sensitive to the effects of heavy prenatal alcohol exposure. Similar adaptive dysfunction is common in other developmental disorders including attention-deficit/hyperactivity disorder (ADHD). ADHD is frequently present in alcohol-exposed children and this overlap in clinical presentation makes identification of alcohol-exposed children difficult. Direct comparison of children with prenatal alcohol exposure and ADHD may yield distinct patterns of cognitive and behavioral performance and add to growing knowledge of the neuropsychological and behavioral profile of prenatal alcohol exposure. Therefore, the aim of the current study was to compare adaptive behavior in children with histories of heavy prenatal alcohol exposure (ALC), nonexposed children with ADHD (ADHD), and typically developing controls (CON). Methods:, Sixty-five children (ALC = 22, ADHD = 23, CON = 20) were selected from a larger ongoing study of the behavioral teratogenicity of alcohol. Alcohol-exposed and control participants were selected to match the ADHD subjects on age, sex, socioeconomic status, and race/ethnicity. Caregivers were administered the Vineland Adaptive Behavior Scales, a semi-structured interview, and were asked to rate their child's behavior on 3 domains of adaptive function. Data were analyzed using regression techniques. Results:, Relative to controls, children in both the ALC and ADHD groups showed adaptive behavior deficits on all 3 domains and children in the ALC group were significantly more impaired than the ADHD group on the daily living skills domain. Within the ALC group, socialization standard scores were lower at older ages. This negative relationship between age and standard scores in the ALC group was also observed on the communication domain, a finding not previously reported. Conclusions:, This study suggests that both children with prenatal alcohol exposure and children with ADHD show impairments in adaptive function relative to controls, but that the pattern of impairment differs between these clinical groups. Adaptive ability in children with prenatal alcohol exposure is characterized by an arrest in development, as evidenced by a lack of improvement with age in socialization and communication scores. In contrast, children with ADHD exhibit a developmental delay in adaptive ability as their scores continued to improve with age, albeit not to the level of control children. Continued research focused on elucidating the patterns of deficits that exist in alcohol-exposed children ultimately will lead to improved differential diagnosis and effective interventions. [source]


    An Exploration Into the Internal Dynamics of a School-Based Mental Health Collaboration

    JOURNAL OF SCHOOL HEALTH, Issue 5 2006
    Kenneth Wm.
    School-based collaboratives are one delivery model that satisfies this expanding role. Lacking in the research on this emerging setting, particularly in the realm of mental health, is information that gives insight into the interactions that occur between collaborative partners who may have opposing aims. This study explores 2 outcome measures, timeliness of implementation of the treatment plan and parental satisfaction with the treatment plan and the influence of the professional view of collaborative partners on these outcomes. Archival data were collected from patient charts on 101 Hispanic students, 73.3% of who were male and the sample mean age was 11.3 years. These data were analyzed using multiple regression techniques. Results revealed that time to implementation was impacted by the diagnosis alignment of participating partners, a proxy for professional view, as was parental satisfaction. Satisfaction was not influenced by timeliness, indicating alternative outcome desires. (J Sch Health. 2006;76(5):164-168) [source]


    FLOOD STAGE FORECASTING WITH SUPPORT VECTOR MACHINES,

    JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION, Issue 1 2002
    Shie-Yui Liong
    ABSTRACT: Machine learning techniques are finding more and more applications in the field of forecasting. A novel regression technique, called Support Vector Machine (SVM), based on the statistical learning theory is explored in this study. SVM is based on the principle of Structural Risk Minimization as opposed to the principle of Empirical Risk Minimization espoused by conventional regression techniques. The flood data at Dhaka, Bangladesh, are used in this study to demonstrate the forecasting capabilities of SVM. The result is compared with that of Artificial Neural Network (ANN) based model for one-lead day to seven-lead day forecasting. The improvements in maximum predicted water level errors by SVM over ANN for four-lead day to seven-lead day are 9.6 cm, 22.6 cm, 4.9 cm and 15.7 cm, respectively. The result shows that the prediction accuracy of SVM is at least as good as and in some cases (particularly at higher lead days) actually better than that of ANN, yet it offers advantages over many of the limitations of ANN, for example in arriving at ANN's optimal network architecture and choosing useful training set. Thus, SVM appears to be a very promising prediction tool. [source]


    Real estate and corporate valuation: an asset pricing perspective

    MANAGERIAL AND DECISION ECONOMICS, Issue 7 2001
    Liow Kim Hiang
    Property is a significant asset in the balance sheets of some Singapore industrial/commerce firms and hotel corporations. In this research, we take on the task of examining the relationship between real estate and stock market valuation of these business firms from an asset pricing perspective. Specifically, the real estate sensitivity of ,property-intensive', non-real estate stocks is investigated in both a three-index (market, sector and property) of stock returns and in an arbitrage pricing theory (APT) framework. The APT model is further recast as a multivariate non-linear regression model with across-equation restrictions. Using weekly returns on ,property-intensive' stocks in the period 1989,1998 and three shorter-sample periods, iterated non-linear seemingly regression techniques (ITNSUR) are employed to obtain joint estimates of stock sensitivities and their associated APT risk ,prices'. The ,real estate' sensitivity is found to be systematic and priced in the APT sense of corporations being paid an ex ante premium for bearing property market risk in investing and owning properties in two of the three sample periods (1989,1991, 1992,1994). The empirical results provide some support that property is a factor in corporate valuation, and is broadly consistent with the efficient markets hypothesis. The implications for portfolio and corporate management are examined. Copyright © 2001 John Wiley & Sons, Ltd. [source]


    Simulation of the mean urban heat island using 2D surface parameters: empirical modelling, verification and extension

    METEOROLOGICAL APPLICATIONS, Issue 3 2009
    Bernadett Balázs
    Abstract The spatial distribution of the annual mean urban heat island (UHI) intensity was simulated applying empirical models based on datasets from urban areas of Szeged and Debrecen, using simple and easily determinable urban surface cover variables. These two cities are situated on the Alföld (Great Hungarian Plain) and have similar topographic and climatic conditions. Temperature field measurements were carried out, Landsat satellite images were evaluated, and then one- and multiple variable models were constructed using linear regression techniques. The selected multiple-parameter models were verified using independent datasets from three urban settlements. In order to obtain some impression of the mean UHI patterns in other cities with no temperature measurements available, the better model was extended to urban areas of four other cities situated in geographical environments similar to Szeged and Debrecen. The main shortcoming of typical empirical models, namely that they are often restricted to a specific location, is overcome by the obtained model since it is not entirely site but more region specific, and valid in a large and densely populated area with several settlements. Copyright © 2009 Royal Meteorological Society [source]


    A mathematical programming approach for improving the robustness of least sum of absolute deviations regression

    NAVAL RESEARCH LOGISTICS: AN INTERNATIONAL JOURNAL, Issue 4 2006
    Avi Giloni
    Abstract This paper discusses a novel application of mathematical programming techniques to a regression problem. While least squares regression techniques have been used for a long time, it is known that their robustness properties are not desirable. Specifically, the estimators are known to be too sensitive to data contamination. In this paper we examine regressions based on Least-sum of Absolute Deviations (LAD) and show that the robustness of the estimator can be improved significantly through a judicious choice of weights. The problem of finding optimum weights is formulated as a nonlinear mixed integer program, which is too difficult to solve exactly in general. We demonstrate that our problem is equivalent to a mathematical program with a single functional constraint resembling the knapsack problem and then solve it for a special case. We then generalize this solution to general regression designs. Furthermore, we provide an efficient algorithm to solve the general nonlinear, mixed integer programming problem when the number of predictors is small. We show the efficacy of the weighted LAD estimator using numerical examples. © 2006 Wiley Periodicals, Inc. Naval Research Logistics, 2006 [source]


    Differences in perceived barriers to volunteering to formal organizations: Lack of time versus lack of interest

    NONPROFIT MANAGEMENT & LEADERSHIP, Issue 3 2007
    Richard A. Sundeen
    This article reports the findings of a study of perceived barriers to volunteering to formal organizations by nonvolunteers. We examine the types of barriers, their frequencies, and their relationships with individual-level sociodemographic characteristics of nonvolunteers. Data from a 2001,2002 national survey conducted by the U.S. Bureau of Labor Statistics are analyzed using multiple regression techniques. Contrary to general expectations, only three barriers,lack of time, lack of interest, and ill health,are fairly common. Furthermore, as expected, different social class groups identify different types of barriers. The article concludes with a discussion of the theoretical implications of the findings (for example, how indicators of social and cultural capital affect access to volunteer opportunities) and practical implications concerning recruitment of volunteers. [source]


    Rates of caesarean section: analysis of global, regional and national estimates

    PAEDIATRIC & PERINATAL EPIDEMIOLOGY, Issue 2 2007
    Ana P. Betrán
    Summary Rates of caesarean section are of concern in both developed and developing countries. We set out to estimate the proportion of births by caesarean section (CS) at national, regional and global levels, describe regional and subregional patterns and correlate rates with other reproductive health indicators. We analysed nationally representative data available from surveys or vital registration systems on the proportion of births by CS. We used local non-parametric regression techniques to correlate CS with maternal mortality ratio, infant and neonatal mortality rates, and the proportion of births attended by skilled health personnel. Although very unevenly distributed, 15% of births worldwide occur by CS. Latin America and the Caribbean show the highest rate (29.2%), and Africa shows the lowest (3.5%). In developed countries, the proportion of caesarean births is 21.1% whereas in least developed countries only 2% of deliveries are by CS. The analysis suggests a strong inverse association between CS rates and maternal, infant and neonatal mortality in countries with high mortality levels. There is some suggestion of a direct positive association at lower levels of mortality. CS levels may respond primarily to economic determinants. [source]


    Use of Commercial Sex Workers Among Hispanic Migrants In North Carolina: Implications for the Spread of HIV

    PERSPECTIVES ON SEXUAL AND REPRODUCTIVE HEALTH, Issue 4 2004
    Emilio A. Parrado
    CONTEXT: Rates of HIV and AIDS have risen among U.S. Hispanics and in migrant-sending regions of Mexico and Central America, pointing to a link between migration and HIV. However, little is known about male migrants' sexual risk behaviors, such as the use of commercial sex workers. METHODS: The prevalence and frequency of commercial sex worker use was examined among 442 randomly selected Hispanic migrants in Durham, North Carolina. Logistic and Poisson regression techniques were used to model predictors of commercial sex worker use, and descriptive data on condom use with commercial sex workers were examined. RESULTS: Twenty-eight percent of respondents reported using the services of a commercial sex worker during the previous year; rates reached 46% among single men and 40% among married men living apart from their wives. Men with spouses in Durham were less likely than other men to use commercial sex workers (odds ratio, 0.1). Among men who used commercial sex workers, the frequency of visits declined with greater education (incidence rate ratio, 0.9) and increased with hourly wage (1.1). Frequency and use declined with years of residence, although the results were of borderline significance. Reported rates of condom use with commercial sex workers were high, but were likely to fall if familiarity with a commercial sex worker increased. CONCLUSIONS: Commercial sex workers represent an important potential source of HIV infection. Educational and behavioral interventions that take into account social context and target the most vulnerable migrants are needed to help migrants and their partners avoid HIV infection. [source]


    Medication use and risk of falls

    PHARMACOEPIDEMIOLOGY AND DRUG SAFETY, Issue 2 2002
    C. Ineke Neutel PhD, FACE
    Abstract Purpose Injuries due to falls are an important public health concern, particularly for the elderly, and effective prevention is an ongoing endeavour. The present study has two related objectives: (1) to describe associations between drug use and falls in an institutionalized population, and (2) to identify a high risk subgroup within the larger population. Methods The initial analysis was based on a population of 227 residents who were followed over a 1-year period. Logistic regression techniques were used to estimate odds ratios (ORs) of the association of falls and drug use. The study of potential ,high-risk' groups employed a case,crossover design to estimate the risk of falling associated with starting a new drug course. Results Relatively weak ORs for risk of falling were observed for various drug classes; the highest OR was for benzodiazepines (BZD) at OR=1.8, (unadjusted). Residents taking multiple drugs were at particular risk for falling, e.g. an OR of 6.1, for those using 10+ drugs. The case,crossover analysis indicated that residents starting a new BZD/antipsychotic were at very high risk (OR,=,11.4,) for experiencing a fall. Conclusions Residents who took many different types of medications, as well as residents starting a new BZD/antipsychotics were at greatly increased risk of falling. These are high risk groups where increased monitoring or adjustments to drug regimens could lead to prevention of falls. Copyright © 2002 John Wiley & Sons, Ltd. [source]


    The Consolidation of the White Southern Congressional Vote: The Roles of Ideology and Party Identification

    POLITICS & POLICY, Issue 3 2008
    Kenneth A. Wink
    This article examines the effects of party identification and ideology on white southerners' vote choices in U.S. House races from 1980 to 1994. Using American National Election Studies data, we employ descriptive statistics and a variety of regression techniques to test these relationships. We find party identification was more important in explaining vote choice in the election of 1994 than in previous years, and a majority of white southerners first identified with Republicans in 1994. We also find ideology had an independent effect on party identification for white southerners throughout the time series. We conclude that increasingly class-based, ideologically polarized parties, opposition to President Clinton and his health care plan, the success of the Republicans in framing the election as a national ideological struggle, and race-based redistricting after 1990 created a tendency of conservative white southerners to identify with Republicans and to vote for Republican House candidates in 1994. [source]


    Is gateway city clustering behind Canada's declining immigrant home-ownership rates?

    POPULATION, SPACE AND PLACE (PREVIOUSLY:-INT JOURNAL OF POPULATION GEOGRAPHY), Issue 5 2009
    Michael Haan
    Abstract Recently, home-ownership rates have been dropping for Canadian immigrants. These declines, although substantial in their own right, are particularly striking when read alongside the trends of the Canadian-born, who've experienced a comparative surge in recent years. Given that immigrants overwhelmingly cluster in Canada's ,gateway' census metropolitan areas (CMAs) of Toronto, Montréal and Vancouver, this paper uses 2-stage least squares regression techniques and the 2001 Census of Canada to identify whether a shift out of Canada's gateway CMAs would improve immigrant home-ownership rates. The results show that although home-ownership propensities are higher for those who live outside of gateway cities, these differences dissolve once the endogeneity of location choice is considered. This suggests that relocation policies are unlikely to yield dramatic national gains. Copyright © 2009 John Wiley & Sons, Ltd. [source]


    Determinants of Antenatal Care Utilization in Three Rural Areas of Vietnam

    PUBLIC HEALTH NURSING, Issue 4 2007
    Lieu Thi Thuy Trinh
    ABSTRACT Objective: To identify factors associated with any use of antenatal care (ANC), gestational age at entry to ANC, number of visits, and overall ANC utilization in the three provinces of Long an, Ben tre, and Quang ngai. Data: The Vietnam-Australia Primary Health Care Project conducted cross-sectional surveys in 1998,1999. Data from 1,335 eligible women were available for analysis. Methods: Explanatory variables were selected using the Andersen Health Seeking Behavior Model and analyzed using multivariate regression techniques. Results: External environment, predisposing characteristics, and need, which existed before contact with ANC providers, were most related to using any ANC and gestational age at entry to ANC. However, ANC services provided to women and personal health care during pregnancy, which could have resulted from initial contact with ANC providers, were most related to continuation of ANC visits and overall ANC utilization. Significant variability in the use of ANC existed between provinces and between subsets of women in each province. Conclusion: To have more women who attend ANC and attend early, promotion should be targeted at high-risk groups of women. However, to improve the number of ANC visits and overall utilization, the quality of ANC services should be improved. [source]