Assessing continuous bivariate effects among different groups through nonparametric regression models: An application to breast cancer detection

In many applications, the joint effect of two continuous covariates on the target binary response may vary across groups defined by levels of a given factor. A testing procedure that would enable this type of surface-by-factor interactions to be detected has been designed. To accomplish this goal, a logistic generalized additive model (GAM) with bivariate continuous interactions varying across groups defined by levels of a factor is considered. A local scoring algorithm based on local linear kernel smoothers was implemented to estimate the proposed logistic GAM. Bootstrap resampling techniques were used for the purpose of testing for factor-by-surface interactions. Given the high computational cost involved, binning techniques were used to speed up computation in the estimation and testing processes. The adequacy of the bootstrap-based test was assessed by means of a simulation study. If a factor-by-surface interaction is detected in the model, it is then established that the use of the odds-ratio curves is very useful in obtaining a direct interpretation of the fitted model. The benefits of using this methodology when analyzing real data are illustrated by applying the technique to the outputs produced by a computerized system dedicated to the early detection of breast cancer.     

Strategic appraisal of energy models for Mozambique

The needs that an energy supply system must meet are constantly changing, due to technological, social and political reasons. Effective energy planning is a dynamic process that is repeated periodically and adjusts to changing conditions. Energy decision makers and planners are no longer able to rely on inductive decision making since they have to investigate the effect of various decision parameters and possible future changes. To help in this process, models have been developed where estimates of future load growth, candidate power plants, fuels and other key factors can be introduced, from which the planners can evaluate decision parameters and the available alternatives. The paper presents the different methodologies and practices that are used by 11 energy models for energy demand forecasting, supply side management and generation expansion planning, demand side management and integrated resource planning. The paper concludes to the presentation of a strategic appraisal of the examined energy models appropriate for energy planning in Mozambique. Three models are proposed for conducting demand forecasting, generation expansion planning and demand side management. Copyright © 2002 John Wiley & Sons, Ltd.     

Significance of fluid regime in biological filtration

A dispersed flow model previously developed to study substrate utilization in unsaturated media was experimentally verified and its practical application was considered. For this purpose, measurements were made using tap water and a synthetic feed solution. The importance of the change in fluid regime as regards to the simultaneous transport and reaction within biological filters were demonstrated. The effect of drop formation and the breakage of liquid jets inside the filters on substrate utilization was also shown.     

Effect of Channel Changes in the St. Clair River During the Present Century

Periodic man-made changes in the outlet of Lake Huron through the St. Clair River date back to the middle of the last century. These artificial channel changes have been well documented during the present century. They consist of dredging for commercial gravel removal in the upper river during 1908–25 and uncompensated navigation improvements for the 7.6-m (25-ft) and 8.2-m (27-ft) projects completed in 1933 and 1962, respectively. The total effect of these changes on the levels of Lakes Michigan and Huron (hydraulically one lake) and on the upper St. Clair River profile was determined with dynamic flow models. The ultimate effect of the above dredging was a permanent lowering of the Lake Michigan-Huron levels 0.27 m (0.89 ft), which represents a tremendous loss of freshwater resource [32 km³ (7.7 mi³)].     

From hospital to municipal cogeneration systems: an Italian case study

A mixed integer linear programming model combined with a more traditional design by scenarios is proposed to optimize facilities size and operation mode of a municipal energy system involving significant civil centres and a hospital. Moving from the need of a new heat and power station for the local hospital due to the construction of new pavilions, the opportunity of involving other centres in the neighbourhood in a distributed cogeneration system is analysed, increasing system complexity step by step. Smaller cogeneration units tailored to hospital needs are rewarding ventures with relatively low risks but, in a country whose traditional power generation systems heavily rely on fossil fuels and where energy policy and market conditions can make it profitable to sell surplus power, district heating systems foster the installation of larger cogenerators and lead thereby to higher profits and to better performance as for primary energy savings and greenhouse gases emission reduction. Copyright © 2006 John Wiley & Sons, Ltd.     

Geometry-Dependent R-Curve for Quasi-Brittle Materials

Fracture of quasi-brittle materials such as ceramics and cement-based materials can be described by an R -curve, defined as fracture resistance. The R -curve can be constructed as the envelope of the strain energy release rates ( G ). In this study, R -curve is defined as an envelope of G -curves obtained by varying the dimensions of identical specimens, while keeping the geometry and the initial flaw size constant. By approximating the G -curve with a second-order function of the crack length, a simple formulation of the R -curve is derived by solving a differential equation. Three parameters are needed for the proposed R -curve. These parameters can be obtained by testing a notched-beam specimen and using linear elastic fracture mechanics. The proposed R -curve simulates well the geometry dependency as well as other characteristics of the fracture response of quasi-brittle materials.     

Ecophysiology of ochratoxigenic Aspergillus ochraceus and Penicillium verrucosum isolates. Predictive models for fungal spoilage prevention - a review

Ochratoxin A (OTA) is a secondary metabolite produced by several species of Aspergillus and Penicillium; among them Aspergillus ochraceus and Penicillium verrucosum are two ochratoxigenic species capable of growing in different climates and thus contamination of food crops with OTA can occur worldwide. OTA can be found in a wide range of foods such as cereals, coffee, cocoa, spices, beer, wine, dried vine fruit, grapes and meat products. OTA is toxic to animals, it presents neurotoxic, immunotoxic and nephrotoxic effects. It has been implicated in a human kidney disorder known as Balkan Endemic Nephropathy. This review focuses on the ecophysiology of ochratoxin-producing Aspergillus ochraceus and Penicillium verrucosum, the effect of environmental factors on their germination, mycelial growth, and OTA production. Knowledge of environmental conditions required for sucessive stages of fungal development represent the first step towards preventing mycotoxin formation. Predictive models for different stages of fungal development are presented, which allow prediction of the time before spoilage as a function of the abiotic factors. Finally, the implications of these studies in management of barley, coffee and grapes are described. This can help to identify the critical control points in their production, storage and distribution processes.