Minimum Hellinger distance based inference for scalar skew-normal and skew-t distributions

The skew-normal is a parametric model that extends the normal family by the addition of a shape parameter to account for skewness. As well, the skew-t distribution is generated by a perturbation of symmetry of the basic Student’s t density. These families share some nice properties. In particular, they allow a continuous variation through different degrees of asymmetry and, in the case of the skew-t, tail thickness, but still retain relevant features of the perturbed symmetric densities. In both models, a problem occurs in the estimation of the skewness parameter: for small and moderate sample sizes, the maximum likelihood method gives rise to an infinite estimate with positive probability, even when the sample skewness is not too large. To get around this phenomenon, we consider the minimum Hellinger distance estimation technique as an alternative to maximum likelihood. The method always leads to a finite estimate of the shape parameter. Furthermore, the procedure is asymptotically efficient under the assumed model and allows for testing hypothesis and setting confidence regions in a standard fashion.     

Absolute measurement of planarity with Fritz's method: uncertainty evaluation

Fritz's method [Opt. Eng. 23, 379 (1984)] of using Zernike polynomials to assess the absolute planarity of test plates is revisited. A refinement is described that takes into account the data decorrelation that appears in experiments. An uncertainty balance is defined by propagation of error contributions through the steps of the method. The resultant measuring procedure is demonstrated on a data set from experiments, and a nanometer level of uncertainty is achieved.     

Interaction between low-profile agents and glass fibres in polyester-based composites

Scanning electron microscope and differential scanning calorimetry studies were used to study the morphology and thermal properties of cured systems constituted of different unsaturated polyester formulations, two conventional low-shrink additives and one suitably synthetized additive. The aim was to clarify the influence of these low-profile agents on the resulting phase structure and on the thermal behaviour of the systems. Moreover, the influence was investigated of low-profile agents on the fibre/matrix adhesion in composite materials reinforced with glass fibres coated by silane coupling agents.     

Influence of chronic alcohol abuse on body weight and energy metabolism: is excess ethanol consumption a risk factor for obesity or malnutrition?

OBJECTIVES: To evaluate the influence of chronic alcohol abuse on body composition and energy metabolism in patients affected by chronic alcoholism (group A) compared with a group of healthy social drinkers (group B). SETTING: A university hospital clinic in Italy. SUBJECTS: A total of 32 alcoholics without clinical or laboratory signs of liver cirrhosis and malabsorption. MEASUREMENTS: Body composition was assessed by anthropometric measurements. Resting energy expenditure (REE) and substrate oxidation rate was measured by indirect calorimetry. Daily caloric intake was computed on the basis of a food diary compiled over 7 days. RESULTS: Alcoholics showed a significantly lower body weight (P < 0.05) and a significant lower fat mass (P < 0.05) compared with controls. A higher waist-to-hip ratio was found in group A than in group B, both as a whole group (P < 0.01) or separated by gender (females, P < 0.01) and males, P < 0.001), indicating a prevalence of fat distribution in the abdominal region in alcoholics. REE was significantly higher in group A than in group B (P < 0.05). The non-protein respiratory quotient was significantly lower in group A than in group B (P < 0.001) with a consequent higher utilization of lipids (P < 0.01) and a lower carbohydrate oxidation (P < 0.05) in group A. The energy intake provided only by food ingestion was found to be significantly higher in group B (P < 0.01), whilst the total caloric intake, computed as food intake plus alcohol intake, was higher in group A (P < 0.01). CONCLUSIONS: Alcoholics, as compared with social drinkers, showed a lower body weight due essentially to a fat mass reduction, a higher REE value normalized by fat-free mass, and a preferential utilization of lipids as energy substrate. These findings might suggest that chronic ethanol abuse is able to determine an impairment of nutritional status due, at least in part, to an alteration of the substrate oxidation.     

Investigation of work-related disorders in truck drivers using RULA method

A high incidence of spinal disorders is observed in professional drivers; in particular, back and neck pain result in high rates of morbidity and low retirement age. A sample of 77 drivers, of rubbish-collection vehicles who sit in a standard posture and of road-washing vehicles, who drive with the neck and trunk flexed, bent and twisted, was studied using RULA, a method for the evaluation of the exposure to risk factors associated with work-related upper-limb disorders. Results showed a significant association between trunk and neck scores and all self-reported pains, aches or discomforts in the trunk or neck regions in all subjects. In particular, the neck score was significant in both postures, reflecting high loading of the neck. Significantly different posture scores were also recorded for drivers using an adjustable vs. a non-adjustable seat.In this first RULA study of the working posture of professional truck drivers, the method proved to be a suitable tool for the rapid evaluation of the loading of neck and trunk.     

Engineered articular cartilage: influence of the scaffold on cell phenotype and proliferation

Articular cartilage defects do not heal. Biodegradable scaffolds have been studied for cartilage engineering in order to implant autologous chondrocytes and help cartilage repair. We tested some new collagen matrices differing in collagen type, origin, structure and methods of extraction and purification, and compared the behavior of human chondrocytes cultured on them. Human chondrocytes were grown for three weeks on four different equine type I collagen matrices, one type I, III porcine collagen matrix and one porcine type II collagen matrix. After 21 days, samples were subjected to histochemical, immunohistochemical and histomorphometric analysis to study phenotype expression and cell adhesion. At 7, 14 and 21 days cell proliferation was studied by incorporation of [3H]-thymidine. Our data evidence that the collagen type influences cell morphology, adhesion and growth; indeed, cellularity and rate of proliferation were significantly higher and cells were rounder on the collagen II matrix than on either of the collagen I matrices. Among the collagen I matrices, we observed a great variability in terms of cell adhesion and proliferation. The present study allowed us to identify one type I collagen matrix and one type II collagen matrix that could be usefully employed as a scaffold for chondrocyte transplantation.     

A plug-in approach to sparse and robust principal component analysis

We propose a method for sparse and robust principal component analysis. The methodology is structured in two steps: first, a robust estimate of the covariance matrix is obtained, then this estimate is plugged-in into an elastic-net regression which enforces sparseness. Our approach provides an intuitive, general and flexible extension of sparse principal component analysis to the robust setting. We also show how to implement the algorithm when the dimensionality exceeds the number of observations by adapting the approach to the use of robust loadings from ROBPCA. The proposed technique is seen to compare well for simulated and real datasets.     

Prediction of lithogenic character of human bile by a quaternary system

The lithogenic character of human gallbladder bile was investigated by a mathematical model on the basis of a clinical data set. Each bile was described by means of the quaternary system bile salts--phospholipids--free fatty acids--cholesterol. Reporting a set of bile samples on the tetrahedron of concentrations, a clear separation emerged between the control bile, the bile from patients with gallstones, and the bile of subjects with gallbladder dyskinesia. This group was characterized by a largely increased content of free fatty acids.     

Robust ordinal regression in preference learning and ranking

Multiple Criteria Decision Aiding (MCDA) offers a diversity of approaches designed for providing the decision maker (DM) with a recommendation concerning a set of alternatives (items, actions) evaluated from multiple points of view, called criteria. This paper aims at drawing attention of the Machine Learning (ML) community upon recent advances in a representative MCDA methodology, called Robust Ordinal Regression (ROR). ROR learns by examples in order to rank a set of alternatives, thus considering a similar problem as Preference Learning (ML-PL) does. However, ROR implements the interactive preference construction paradigm, which should be perceived as a mutual learning of the model and the DM. The paper clarifies the specific interpretation of the concept of preference learning adopted in ROR and MCDA, comparing it to the usual concept of preference learning considered within ML. This comparison concerns a structure of the considered problem, types of admitted preference information, a character of the employed preference models, ways of exploiting them, and techniques to arrive at a final ranking.