Learning wind fields with multiple kernels

This paper presents multiple kernel learning (MKL) regression as an exploratory spatial data analysis and modelling tool. The MKL approach is introduced as an extension of support vector regression, where MKL uses dedicated kernels to divide a given task into sub-problems and to treat them separately in an effective way. It provides better interpretability to non-linear robust kernel regression at the cost of a more complex numerical optimization. In particular, we investigate the use of MKL as a tool that allows us to avoid using ad-hoc topographic indices as covariables in statistical models in complex terrains. Instead, MKL learns these relationships from the data in a non-parametric fashion. A study on data simulated from real terrain features confirms the ability of MKL to enhance the interpretability of data-driven models and to aid feature selection without degrading predictive performances. Here we examine the stability of the MKL algorithm with respect to the number of training data samples and to the presence of noise. The results of a real case study are also presented, where MKL is able to exploit a large set of terrain features computed at multiple spatial scales, when predicting mean wind speed in an Alpine region.

Monitoring the clinical introduction of a glutamine and antioxidant solution in critically ill trauma and burn patients

OBJECTIVE: Enteral glutamine supplementation and antioxidants have been shown to be beneficial in some categories of critically ill patients. This study investigated the impact on organ function and clinical outcome of an enteral solution enriched with glutamine and antioxidant micronutrients in patients with trauma and with burns. METHODS: This was a prospective study of a historical control group including critically ill, burned and major trauma patients (n = 86, 40 patients with burns and 46 with trauma, 43 in each group) on admission to an intensive care unit in a university hospital (matching for severity, age, and sex). The intervention aimed to deliver a 500-mL enteral solution containing 30 g of glutamine per day, selenium, zinc, and vitamin E (Gln-AOX) for a maximum of 10 d, in addition to control treatment consisting of enteral nutrition in all patients and intravenous trace elements in all burn patients. RESULTS: Patients were comparable at baseline, except for more inhalation injuries in the burn-Gln-AOX group (P = 0.10) and greater neurologic impairment in the trauma-Gln-AOX group (P = 0.022). Intestinal tolerance was good. The full 500-mL dose was rarely delivered, resulting in a low mean glutamine daily dose (22 g for burn patients and 16 g for trauma patients). In burn patients intravenous trace element delivery was superior to the enteral dose. The evolution of the Sequential Organ Failure Assessment score and other outcome variables did not differ significantly between groups. C-reactive protein decreased faster in the Gln-AOX group. CONCLUSION: The Gln-AOX supplement was well tolerated in critically ill, injured patients, but did not improve outcome significantly. The delivery of glutamine below the 0.5-g/kg recommended dose in association with high intravenous trace element substitution doses in burn patients are likely to have blunted the impact by not reaching an efficient treatment dose. Further trials testing higher doses of Gln are required.

Integration of local-scale hydrological and regional-scale geophysical based on a nonlinear Bayesian sequential simulation approach

Significant progress has been made with regard to the quantitative integration of geophysical and hydrological data at the local scale. However, extending the corresponding approaches to the regional scale represents a major, and as-of-yet largely unresolved, challenge. To address this problem, we have developed a downscaling procedure based on a non-linear Bayesian sequential simulation approach. The basic objective of this algorithm is to estimate the value of the sparsely sampled hydraulic conductivity at non-sampled locations based on its relation to the electrical conductivity, which is available throughout the model space. The in situ relationship between the hydraulic and electrical conductivities is described through a non-parametric multivariate kernel density function. This method is then applied to the stochastic integration of low-resolution, re- gional-scale electrical resistivity tomography (ERT) data in combination with high-resolution, local-scale downhole measurements of the hydraulic and electrical conductivities. Finally, the overall viability of this downscaling approach is tested and verified by performing and comparing flow and transport simulation through the original and the downscaled hydraulic conductivity fields. Our results indicate that the proposed procedure does indeed allow for obtaining remarkably faithful estimates of the regional-scale hydraulic conductivity structure and correspondingly reliable predictions of the transport characteristics over relatively long distances.

Strengthening of Existing Single Span Steel Beam and Concrete Deck Bridges, HR-238, Part 1, 1983

The Phase I research, Iowa Department of Transportation (IDOT) Project HR-214, "Feasibility Study of Strengthening Existing Single Span Steel Beam Concrete Deck Bridges," verified that post-tensioning can be used to provide strengthening of the composite bridges under investigation. Phase II research, reported here, involved the strengthening of two full-scale prototype bridges - one a prototype of the model bridge tested during Phase I and the other larger and skewed. In addition to the field work, Phase II also involved a considerable amount of laboratory work. A literature search revealed that only minimal data existed on the angle-plus-bar shear connectors. Thus, several specimens utilizing angle-plus-bar, as well as channels, studs and high strength bolts as shear connectors were fabricated and tested. To obtain additional shear connector information, the bridge model of Phase I was sawed into four composite concrete slab and steel beam specimens. Two of the resulting specimens were tested with the original shear connection, while the other two specimens had additional shear connectors added before testing. Although orthotropic plate theory was shown in Phase I to predict vertical load distribution in bridge decks and to predict approximate distribution of post-tensioning for right-angle bridges, it was questioned whether the theory could also be used on skewed bridges. Thus, a small plexiglas model was constructed and used in vertical load distribution tests and post-tensioning force distribution tests for verification of the theory. Conclusions of this research are as follows: (1) The capacity of existing shear connectors must be checked as part of a bridge strengthening program. Determination of the concrete deck strength in advance of bridge strengthening is also recommended. (2) The ultimate capacity of angle-plus-bar shear connectors can be computed on the basis of a modified AASHTO channel connector formula and an angle-to-beam weld capacity check. (3) Existing shear connector capacity can be augmented by means of double-nut high strength bolt connectors. (4) Post-tensioning did not significantly affect truck load distribution for right angle or skewed bridges. (5) Approximate post-tensioning and truck load distribution for actual bridges can be predicted by orthotropic plate theory for vertical load; however, the agreement between actual distribution and theoretical distribution is not as close as that measured for the laboratory model in Phase I. (6) The right angle bridge exhibited considerable end restraint at what would be assumed to be simple support. The construction details at bridge abutments seem to be the reason for the restraint. (7) The skewed bridge exhibited more end restraint than the right angle bridge. Both skew effects and construction details at the abutments accounted for the restraint. (8) End restraint in the right angle and skewed bridges reduced tension strains in the steel bridge beams due to truck loading, but also reduced the compression strains caused by post-tensioning.

Aggregation of liposomes induced by calcium: A structural and kinetic study

In this work, the calcium-induced aggregation of phosphatidylserine liposomes is probed by means of the analysis of the kinetics of such process as well as the aggregate morphology. This novel characterization of liposome aggregation involves the use of static and dynamic light-scattering techniques to obtain kinetic exponents and fractal dimensions. For salt concentrations larger than 5 mM, a diffusion-limited aggregation regime is observed and the Brownian kernel properly describes the time evolution of the diffusion coefficient. For slow kinetics, a slightly modified multiple contact kernel is required. In any case, a time evolution model based on the numerical resolution of Smoluchowski's equation is proposed in order to establish a theoretical description for the aggregating system. Such a model provides an alternative procedure to determine the dimerization constant, which might supply valuable information about interaction mechanisms between phospholipid vesicles.

Deep learning via semi-supervised embedding

We show how nonlinear embedding algorithms popular for use with shallow semi-supervised learning techniques such as kernel methods can be applied to deep multilayer architectures, either as a regularizer at the output layer, or on each layer of the architecture. This provides a simple alternative to existing approaches to deep learning whilst yielding competitive error rates compared to those methods, and existing shallow semi-supervised techniques.

Statistical assessment of dataset shift and model portability in multi-angle in-track image acquisitions

In this study we propose an evaluation of the angular effects altering the spectral response of the land-cover over multi-angle remote sensing image acquisitions. The shift in the statistical distribution of the pixels observed in an in-track sequence of WorldView-2 images is analyzed by means of a kernel-based measure of distance between probability distributions. Afterwards, the portability of supervised classifiers across the sequence is investigated by looking at the evolution of the classification accuracy with respect to the changing observation angle. In this context, the efficiency of various physically and statistically based preprocessing methods in obtaining angle-invariant data spaces is compared and possible synergies are discussed.

Predictive analysis and mapping of indoor radon concentrations in Switzerland

It is estimated that around 230 people die each year due to radon (222Rn) exposure in Switzerland. 222Rn occurs mainly in closed environments like buildings and originates primarily from the subjacent ground. Therefore it depends strongly on geology and shows substantial regional variations. Correct identification of these regional variations would lead to substantial reduction of 222Rn exposure of the population based on appropriate construction of new and mitigation of already existing buildings. Prediction of indoor 222Rn concentrations (IRC) and identification of 222Rn prone areas is however difficult since IRC depend on a variety of different variables like building characteristics, meteorology, geology and anthropogenic factors. The present work aims at the development of predictive models and the understanding of IRC in Switzerland, taking into account a maximum of information in order to minimize the prediction uncertainty. The predictive maps will be used as a decision-support tool for 222Rn risk management. The construction of these models is based on different data-driven statistical methods, in combination with geographical information systems (GIS). In a first phase we performed univariate analysis of IRC for different variables, namely the detector type, building category, foundation, year of construction, the average outdoor temperature during measurement, altitude and lithology. All variables showed significant associations to IRC. Buildings constructed after 1900 showed significantly lower IRC compared to earlier constructions. We observed a further drop of IRC after 1970. In addition to that, we found an association of IRC with altitude. With regard to lithology, we observed the lowest IRC in sedimentary rocks (excluding carbonates) and sediments and the highest IRC in the Jura carbonates and igneous rock. The IRC data was systematically analyzed for potential bias due to spatially unbalanced sampling of measurements. In order to facilitate the modeling and the interpretation of the influence of geology on IRC, we developed an algorithm based on k-medoids clustering which permits to define coherent geological classes in terms of IRC. We performed a soil gas 222Rn concentration (SRC) measurement campaign in order to determine the predictive power of SRC with respect to IRC. We found that the use of SRC is limited for IRC prediction. The second part of the project was dedicated to predictive mapping of IRC using models which take into account the multidimensionality of the process of 222Rn entry into buildings. We used kernel regression and ensemble regression tree for this purpose. We could explain up to 33% of the variance of the log transformed IRC all over Switzerland. This is a good performance compared to former attempts of IRC modeling in Switzerland. As predictor variables we considered geographical coordinates, altitude, outdoor temperature, building type, foundation, year of construction and detector type. Ensemble regression trees like random forests allow to determine the role of each IRC predictor in a multidimensional setting. We found spatial information like geology, altitude and coordinates to have stronger influences on IRC than building related variables like foundation type, building type and year of construction. Based on kernel estimation we developed an approach to determine the local probability of IRC to exceed 300 Bq/m3. In addition to that we developed a confidence index in order to provide an estimate of uncertainty of the map. All methods allow an easy creation of tailor-made maps for different building characteristics. Our work is an essential step towards a 222Rn risk assessment which accounts at the same time for different architectural situations as well as geological and geographical conditions. For the communication of 222Rn hazard to the population we recommend to make use of the probability map based on kernel estimation. The communication of 222Rn hazard could for example be implemented via a web interface where the users specify the characteristics and coordinates of their home in order to obtain the probability to be above a given IRC with a corresponding index of confidence. Taking into account the health effects of 222Rn, our results have the potential to substantially improve the estimation of the effective dose from 222Rn delivered to the Swiss population.

Prototipo de RINA sobre Ethernet

Este PFM se enmarca dentro del trabajo a realizar por el proyecto IRATI. IRATI es un proyecto STReP (Specific Targeted Research Project) financiado por la Unión Europea dentro del programa FP7 (Seventh Framework Programme for Research and Technological Development). El objetivo general de IRATI es conseguir una mayor compresión y exploración de RINA. El trabajo que se reportará en este PFM es (aproximadamente) la primera fase del diseño y desarrollo de un prototipo de RINA sobre Ethernet en el seno del Kernel (Linux), basándose y generando software libre.

Update on the non-prewhitening model observer in computed tomography for the assessment of the adaptive statistical and model-based iterative reconstruction algorithms.

The state of the art to describe image quality in medical imaging is to assess the performance of an observer conducting a task of clinical interest. This can be done by using a model observer leading to a figure of merit such as the signal-to-noise ratio (SNR). Using the non-prewhitening (NPW) model observer, we objectively characterised the evolution of its figure of merit in various acquisition conditions. The NPW model observer usually requires the use of the modulation transfer function (MTF) as well as noise power spectra. However, although the computation of the MTF poses no problem when dealing with the traditional filtered back-projection (FBP) algorithm, this is not the case when using iterative reconstruction (IR) algorithms, such as adaptive statistical iterative reconstruction (ASIR) or model-based iterative reconstruction (MBIR). Given that the target transfer function (TTF) had already shown it could accurately express the system resolution even with non-linear algorithms, we decided to tune the NPW model observer, replacing the standard MTF by the TTF. It was estimated using a custom-made phantom containing cylindrical inserts surrounded by water. The contrast differences between the inserts and water were plotted for each acquisition condition. Then, mathematical transformations were performed leading to the TTF. As expected, the first results showed a dependency of the image contrast and noise levels on the TTF for both ASIR and MBIR. Moreover, FBP also proved to be dependent of the contrast and noise when using the lung kernel. Those results were then introduced in the NPW model observer. We observed an enhancement of SNR every time we switched from FBP to ASIR to MBIR. IR algorithms greatly improve image quality, especially in low-dose conditions. Based on our results, the use of MBIR could lead to further dose reduction in several clinical applications.

Sedentarism affects body fat mass index and fat-free mass index in adults aged 18 to 98 years.

OBJECTIVE: Body mass index does not discriminate body fat from fat-free mass or determine changes in these parameters with physical activity and aging. Body fat mass index (BFMI) and fat-free mass index (FFMI) permit comparisons of subjects with different heights. This study evaluated differences in body mass index, BFMI, and FFMI in physically active and sedentary subjects younger and older than 60 y and determined the association between physical activity, age, and body composition parameters in a healthy white population between ages 18 and 98 y. METHODS: Body fat and fat-free mass were determined in healthy white men (n = 3549) and women (n = 3184), between ages 18 and 98 y, by bioelectrical impedance analysis. BFMI and FFMI (kg/m2) were calculated. Physical activity was defined as at least 3 h/wk of endurance-type activity for at least 2 mo. RESULTS: Physically active as opposed to sedentary subjects were more likely to have a low BFMI (men: odds ratio [OR], 1.4; confidence interval [CI], 0.7-2.5; women: OR 1.9, CI 1.6-2.2) and less likely to have very high BFMI (men: OR, 0.2; CI, 0.1-0.2; women: OR, 0.1; CI, 0.02-0.2), low FFMI (men: OR, 0.5; CI, 0.3-0.9; women: OR, 0.7; CI, 0.6-0.9), or very high FFMI (men: OR, 0.6; CI, 0.4-0.8; women: OR, 0.7; CI, 0.5-1.0). Compared with subjects younger than 60 y, those older than 60 y were more like to have very high BFMI (men: OR, 6.5; CI, 4.5-9.3; women: OR, 14.0; CI, 9.6-20.5), and women 60 y and older were less likely to have a low BFMI (OR, 0.4; CI, 0.2-0.5). CONCLUSIONS: A clear association was found between low physical activity or age and height-normalized body composition parameters (BFMI and FFMI) derived from bioelectrical impedance analysis. Physically active subjects were more likely to have high or very high or low FFMI. Older subjects had higher body weights and BFMI.

Multi-source composite kernels for urban image classification

This letter presents advanced classification methods for very high resolution images. Efficient multisource information, both spectral and spatial, is exploited through the use of composite kernels in support vector machines. Weighted summations of kernels accounting for separate sources of spectral and spatial information are analyzed and compared to classical approaches such as pure spectral classification or stacked approaches using all the features in a single vector. Model selection problems are addressed, as well as the importance of the different kernels in the weighted summation.

Axiomatization of the nucl eolus of assignment games

On the domain of general assignment games (with possible reservation prices) the core is axiomatized as the unique solution satisfying two consistency principles: projection consistency and derived consistency. Also, an axiomatic characterization of the nucleolus is given as the unique solution that satisfies derived consistency and equal maximum complaint between groups. As a consequence, we obtain a geometric characterization of the nucleolus. Maschler et al. (1979) provide a geometrical characterization for the intersection of the kernel and the core of a coalitional game, showing that those allocations that lie in both sets are always the midpoint of certain bargaining range between each pair of players. In the case of the assignment game, this means that the kernel can be determined as those core allocations where the maximum amount, that can be transferred without getting outside the core, from one agent to his / her optimally matched partner equals the maximum amount that he / she can receive from this partner, also remaining inside the core. We now prove that the nucleolus of the assignment game can be characterized by requiring this bisection property be satisfied not only for optimally matched pairs but also for optimally matched coalitions. Key words: cooperative games, assignment game, core, nucleolus

Mixed spectral-structural classification of very high resolution images with summation kernels

In this paper, mixed spectral-structural kernel machines are proposed for the classification of very-high resolution images. The simultaneous use of multispectral and structural features (computed using morphological filters) allows a significant increase in classification accuracy of remote sensing images. Subsequently, weighted summation kernel support vector machines are proposed and applied in order to take into account the multiscale nature of the scene considered. Such classifiers use the Mercer property of kernel matrices to compute a new kernel matrix accounting simultaneously for two scale parameters. Tests on a Zurich QuickBird image show the relevance of the proposed method : using the mixed spectral-structural features, the classification accuracy increases of about 5%, achieving a Kappa index of 0.97. The multikernel approach proposed provide an overall accuracy of 98.90% with related Kappa index of 0.985.

Fructose and metabolic diseases: new findings, new questions.

There has been much concern regarding the role of dietary fructose in the development of metabolic diseases. This concern arises from the continuous increase in fructose (and total added caloric sweeteners consumption) in recent decades, and from the increased use of high-fructose corn syrup (HFCS) as a sweetener. A large body of evidence shows that a high-fructose diet leads to the development of obesity, diabetes, and dyslipidemia in rodents. In humans, fructose has long been known to increase plasma triglyceride concentrations. In addition, when ingested in large amounts as part of a hypercaloric diet, it can cause hepatic insulin resistance, increased total and visceral fat mass, and accumulation of ectopic fat in the liver and skeletal muscle. These early effects may be instrumental in causing, in the long run, the development of the metabolic syndrome. There is however only limited evidence that fructose per se, when consumed in moderate amounts, has deleterious effects. Several effects of a high-fructose diet in humans can be observed with high-fat or high-glucose diets as well, suggesting that an excess caloric intake may be the main factor involved in the development of the metabolic syndrome. The major source of fructose in our diet is with sweetened beverages (and with other products in which caloric sweeteners have been added). The progressive replacement of sucrose by HFCS is however unlikely to be directly involved in the epidemy of metabolic disease, because HFCS appears to have basically the same metabolic effects as sucrose. Consumption of sweetened beverages is however clearly associated with excess calorie intake, and an increased risk of diabetes and cardiovascular diseases through an increase in body weight. This has led to the recommendation to limit the daily intake of sugar calories.