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.

How Much Does It Cost? Optimization of Costs in Sequence Analysis of Social Science Data

One major methodological problem in analysis of sequence data is the determination of costs from which distances between sequences are derived. Although this problem is currently not optimally dealt with in the social sciences, it has some similarity with problems that have been solved in bioinformatics for three decades. In this article, the authors propose an optimization of substitution and deletion/insertion costs based on computational methods. The authors provide an empirical way of determining costs for cases, frequent in the social sciences, in which theory does not clearly promote one cost scheme over another. Using three distinct data sets, the authors tested the distances and cluster solutions produced by the new cost scheme in comparison with solutions based on cost schemes associated with other research strategies. The proposed method performs well compared with other cost-setting strategies, while it alleviates the justification problem of cost schemes.

Using register data to assess the impact of response enhancement methods on the risk of nonresponse bias and survey costs

The pursuit of high response rates to minimise the threat of nonresponse bias continues to dominate decisions about resource allocation in survey research. Yet a growing body of research has begun to question this practice. In this study, we use previously unavailable data from a new sampling frame based on population registers to assess the value of different methods designed to increase response rates on the European Social Survey in Switzerland. Using sampling data provides information about both respondents and nonrespondents, making it possible to examine how changes in response rates resulting from the use of different fieldwork methods relate to changes in the composition and representativeness of the responding sample. We compute an R-indicator to assess representativity with respect to the sampling register variables, and find little improvement in the sample composition as response rates increase. We then examine the impact of response rate increases on the risk of nonresponse bias based on Maximal Absolute Bias (MAB), and coefficients of variation between subgroup response rates, alongside the associated costs of different types of fieldwork effort. The results show that increases in response rate help to reduce MAB, while only small but important improvements to sample representativity are gained by varying the type of effort. These findings lend further support to research that has called into question the value of extensive investment in procedures aimed at reaching response rate targets and the need for more tailored fieldwork strategies aimed both at reducing survey costs and minimising the risk of bias.

Multiple Kernel Learning of Environmental Data. Case study: Analysis and Mapping of Wind Fields

The paper presents the Multiple Kernel Learning (MKL) approach as a modelling and data exploratory tool and applies it to the problem of wind speed mapping. Support Vector Regression (SVR) is used to predict spatial variations of the mean wind speed from terrain features (slopes, terrain curvature, directional derivatives) generated at different spatial scales. Multiple Kernel Learning is applied to learn kernels for individual features and thematic feature subsets, both in the context of feature selection and optimal parameters determination. An empirical study on real-life data confirms the usefulness of MKL as a tool that enhances the interpretability of data-driven models.

Optimal dividend strategies for a compound Poisson risk process under transaction costs and power utility

We characterize the value function of maximizing the total discounted utility of dividend payments for a compound Poisson insurance risk model when strictly positive transaction costs are included, leading to an impulse control problem. We illustrate that well known simple strategies can be optimal in the case of exponential claim amounts. Finally we develop a numerical procedure to deal with general claim amount distributions.

Coûts humains de la dépendance au tabac [Human costs of tobacco dependence]

Human costs are an indicator of the value attributed to human burden related to diseases. The present paper summarizes a Swiss study. led by economists and physicians, which estimate the monetary value attributed by the general population regarding the consequences on quality of life of smoking related diseases. Such economic study may contribute to better understand smokers' concerns and beliefs and to improve effective strategies for smoking cessation in primary care.

What can we learn from international comparisons of costs by DRG ?

OBJECTIVES: The objective of this study was to compare costs data by diagnosis related group (DRG) between Belgium and Switzerland. Our hypotheses were that differences between countries can probably be explained by methodological differences in cost calculations, by differences in medical practices and by differences in cost structures within the two countries. METHODS: Classifications of DRG used in the two countries differ (AP-DRGs version 1.7 in Switzerland and APR-DRGs version 15.0 in Belgium). The first step of this study was to transform Belgian summaries into Swiss AP-DRGs. Belgian and Swiss data were calculated with a clinical costing methodology (full costing). Belgian and Swiss costs were converted into US$ PPP (purchasing power parity) in order to neutralize differences in purchasing power between countries. RESULTS: The results of this study showed higher costs in Switzerland despite standardization of cost data according to PPP. The difference is not explained by the case-mix index because this was similar for inliers between the two countries. The length of stay (LOS) was also quite similar for inliers between the two countries. The case-mix index was, however, higher for high outliers in Belgium, as reflected in a higher LOS for these patients. Higher costs in Switzerland are thus probably explained mainly by the higher number of agency staff by service in this country or because of differences in medical practices. CONCLUSIONS: It is possible to make international comparisons but only if there is standardization of the case-mix between countries and only if comparable accountancy methodologies are used. Harmonization of DRGs groups, nomenclature and accountancy is thus required.

The costs of disorders of the brain in Switzerland: an update from the European Brain Council Study for 2010.

BACKGROUND: In 2005, findings of the first "cost of disorders of the brain in Europe" study of the European Brain Council (EBC) showed that these costs cause a substantial economic burden to the Swiss society. In 2010 an improved update with a broader range of disorders has been analysed. This report shows the new findings for Switzerland and discusses changes. METHODS: Data are derived from the EBC 2010 census study that estimates 12-month prevalence of 12 groups of disorders of the brain and calculates costs (direct health-care costs, direct non-medical costs and indirect costs) by combining top-down and bottom up cost approaches using existing data. RESULTS: The most frequent disorder was headache (2.3 million). Anxiety disorders were found in 1 million persons and sleep disorders in 700,000 persons. Annual costs for all assessed disorders total to 14.5 billion Euro corresponding to about 1,900 EUR per inhabitant per year. Mood, psychotic disorders and dementias (appr. 2 billion EUR each) were most costly. Costs per person were highest for neurological/neurosurgery-relevant disorders, e.g. neuromuscular disorders, brain tumour and multiple sclerosis (38,000 to 24,000 EUR). CONCLUSION: The estimates of the EBC 2010 study for Switzerland provide a basis for health care planning. Increase in size and costs compared to 2005 are mostly due to the inclusion of new disorders (e.g., sleep disorders), or the re-definition of others (e.g., headache) and to an increase in younger cohorts. We suggest coordinated research and preventive measures coordinated between governmental bodies, private health-care and pharmaceutical companies.

Does a mandatory telemedicine call prior to visiting a physician reduce costs or simply attract good risks?

This paper aims to estimate empirically the efficiency of a Swiss telemedicine service introduced in 2003. We used claims' data gathered by a major Swiss health insurer, over a period of 6 years and involving 160 000 insured adults. In Switzerland, health insurance is mandatory, but everyone has the option of choosing between a managed care plan and a fee-for-service plan. This paper focuses on a conventional fee-for-service plan including a mandatory access to a telemedicine service; the insured are obliged to phone this medical call centre before visiting a physician. This type of plan generates much lower average health expenditures than a conventional insurance plan. Reasons for this may include selection, incentive effects or efficiency. In our sample, about 90% of the difference in health expenditure can be explained by selection and incentive effects. The remaining 10% of savings due to the efficiency of the telemedicine service amount to about SFr 150 per year per insured, of which approximately 60% is saved by the insurer and 40% by the insured. Although the efficiency effect is greater than the cost of the plan, the big winners are the insured who not only save monetary and non-monetary costs but also benefit from reduced premiums. Copyright © 2010 John Wiley & Sons, Ltd.

Integrating life cycle costs and environmental impacts of composite rail car-bodies for a Korean train

Background, aim, and scope A coupled Life Cycle Costing and life cycle assessment has been performed for car-bodies of the Korean Tilting Train eXpress (TTX) project using European and Korean databases, with the objective of assessing environmental and cost performance to aid materials and process selection. More specifically, the potential of polymer composite car-body structures for the Korean Tilting Train eXpress (TTX) has been investigated. Materials and methods This assessment includes the cost of both carriage manufacturing and use phases, coupled with the life cycle environmental impacts of all stages from raw material production, through carriage manufacture and use, to end-of-life scenarios. Metallic carriages were compared with two composite options: hybrid steel-composite and full-composite carriages. The total planned production for this regional Korean train was 440 cars, with an annual production volume of 80 cars. Results and discussion The coupled analyses were used to generate plots of cost versus energy consumption and environmental impacts. The results show that the raw material and manufacturing phase costs are approximately half of the total life cycle costs, whilst their environmental impact is relatively insignificant (3-8%). The use phase of the car-body has the largest environmental impact for all scenarios, with near negligible contributions from the other phases. Since steel rail carriages weigh more (27-51%), the use phase cost is correspondingly higher, resulting in both the greatest environmental impact and the highest life cycle cost. Compared to the steel scenario, the hybrid composite variant has a lower life cycle cost (16%) and a lower environmental impact (26%). Though the full composite rail carriage may have the highest manufacturing cost, it results in the lowest total life cycle costs and lowest environmental impacts. Conclusions and recommendations This coupled cost and life cycle assessment showed that the full composite variant was the optimum solution. This case study showed that coupling of technical cost models with life cycle assessment offers an efficient route to accurately evaluate economic and environmental performance in a consistent way.

Spatial prediction of monthly wind speeds in complex terrain with adaptive general regression neural networks

This paper presents the general regression neural networks (GRNN) as a nonlinear regression method for the interpolation of monthly wind speeds in complex Alpine orography. GRNN is trained using data coming from Swiss meteorological networks to learn the statistical relationship between topographic features and wind speed. The terrain convexity, slope and exposure are considered by extracting features from the digital elevation model at different spatial scales using specialised convolution filters. A database of gridded monthly wind speeds is then constructed by applying GRNN in prediction mode during the period 1968-2008. This study demonstrates that using topographic features as inputs in GRNN significantly reduces cross-validation errors with respect to low-dimensional models integrating only geographical coordinates and terrain height for the interpolation of wind speed. The spatial predictability of wind speed is found to be lower in summer than in winter due to more complex and weaker wind-topography relationships. The relevance of these relationships is studied using an adaptive version of the GRNN algorithm which allows to select the useful terrain features by eliminating the noisy ones. This research provides a framework for extending the low-dimensional interpolation models to high-dimensional spaces by integrating additional features accounting for the topographic conditions at multiple spatial scales. Copyright (c) 2012 Royal Meteorological Society.