Machine learning algorithms for spatial data. Case studies: Environmental pollution, natural hazards, renewable resources

This paper presents general problems and approaches for the spatial data analysis using machine learning algorithms. Machine learning is a very powerful approach to adaptive data analysis, modelling and visualisation. The key feature of the machine learning algorithms is that they learn from empirical data and can be used in cases when the modelled environmental phenomena are hidden, nonlinear, noisy and highly variable in space and in time. Most of the machines learning algorithms are universal and adaptive modelling tools developed to solve basic problems of learning from data: classification/pattern recognition, regression/mapping and probability density modelling. In the present report some of the widely used machine learning algorithms, namely artificial neural networks (ANN) of different architectures and Support Vector Machines (SVM), are adapted to the problems of the analysis and modelling of geo-spatial data. Machine learning algorithms have an important advantage over traditional models of spatial statistics when problems are considered in a high dimensional geo-feature spaces, when the dimension of space exceeds 5. Such features are usually generated, for example, from digital elevation models, remote sensing images, etc. An important extension of models concerns considering of real space constrains like geomorphology, networks, and other natural structures. Recent developments in semi-supervised learning can improve modelling of environmental phenomena taking into account on geo-manifolds. An important part of the study deals with the analysis of relevant variables and models' inputs. This problem is approached by using different feature selection/feature extraction nonlinear tools. To demonstrate the application of machine learning algorithms several interesting case studies are considered: digital soil mapping using SVM, automatic mapping of soil and water system pollution using ANN; natural hazards risk analysis (avalanches, landslides), assessments of renewable resources (wind fields) with SVM and ANN models, etc. The dimensionality of spaces considered varies from 2 to more than 30. Figures 1, 2, 3 demonstrate some results of the studies and their outputs. Finally, the results of environmental mapping are discussed and compared with traditional models of geostatistics.

Transgenic plant factories for the production of biopolymers and plateform chemicals

In the mid-to long-term, resource constraints will force society to cover a significant share of the future demand for fuels, materials and chemicals by renewable resources. This trend is already visible in the increasing conversion of carbohydrates and plant oils to fuels, chemicals, and polymers. In this perspective, we discuss current efforts and ideas to produce platform chemicals and polymers directly in transgenic plants.

Kernel-based Mapping of Meteorological Fields in Complex Orography

Due to the advances in sensor networks and remote sensing technologies, the acquisition and storage rates of meteorological and climatological data increases every day and ask for novel and efficient processing algorithms. A fundamental problem of data analysis and modeling is the spatial prediction of meteorological variables in complex orography, which serves among others to extended climatological analyses, for the assimilation of data into numerical weather prediction models, for preparing inputs to hydrological models and for real time monitoring and short-term forecasting of weather.In this thesis, a new framework for spatial estimation is proposed by taking advantage of a class of algorithms emerging from the statistical learning theory. Nonparametric kernel-based methods for nonlinear data classification, regression and target detection, known as support vector machines (SVM), are adapted for mapping of meteorological variables in complex orography.With the advent of high resolution digital elevation models, the field of spatial prediction met new horizons. In fact, by exploiting image processing tools along with physical heuristics, an incredible number of terrain features which account for the topographic conditions at multiple spatial scales can be extracted. Such features are highly relevant for the mapping of meteorological variables because they control a considerable part of the spatial variability of meteorological fields in the complex Alpine orography. For instance, patterns of orographic rainfall, wind speed and cold air pools are known to be correlated with particular terrain forms, e.g. convex/concave surfaces and upwind sides of mountain slopes.Kernel-based methods are employed to learn the nonlinear statistical dependence which links the multidimensional space of geographical and topographic explanatory variables to the variable of interest, that is the wind speed as measured at the weather stations or the occurrence of orographic rainfall patterns as extracted from sequences of radar images. Compared to low dimensional models integrating only the geographical coordinates, the proposed framework opens a way to regionalize meteorological variables which are multidimensional in nature and rarely show spatial auto-correlation in the original space making the use of classical geostatistics tangled.The challenges which are explored during the thesis are manifolds. First, the complexity of models is optimized to impose appropriate smoothness properties and reduce the impact of noisy measurements. Secondly, a multiple kernel extension of SVM is considered to select the multiscale features which explain most of the spatial variability of wind speed. Then, SVM target detection methods are implemented to describe the orographic conditions which cause persistent and stationary rainfall patterns. Finally, the optimal splitting of the data is studied to estimate realistic performances and confidence intervals characterizing the uncertainty of predictions.The resulting maps of average wind speeds find applications within renewable resources assessment and opens a route to decrease the temporal scale of analysis to meet hydrological requirements. Furthermore, the maps depicting the susceptibility to orographic rainfall enhancement can be used to improve current radar-based quantitative precipitation estimation and forecasting systems and to generate stochastic ensembles of precipitation fields conditioned upon the orography.

Production of renewable polymers from crop plants

Plants produce a range of biopolymers for purposes such as maintenance of structural integrity, carbon storage, and defense against pathogens and desiccation. Several of these natural polymers are used by humans as food and materials, and increasingly as an energy carrier. In this review, we focus on plant biopolymers that are used as materials in bulk applications, such as plastics and elastomers, in the context of depleting resources and climate change, and consider technical and scientific bottlenecks in the production of novel or improved materials in transgenic or alternative crop plants. The biopolymers discussed are natural rubber and several polymers that are not naturally produced in plants, such as polyhydroxyalkanoates, fibrous proteins and poly-amino acids. In addition, monomers or precursors for the chemical synthesis of biopolymers, such as 4-hydroxybenzoate, itaconic acid, fructose and sorbitol, are discussed briefly

Stress Perception in Public Organizations: Expanding the Job Demands - Job Resources Model by Including Public Service Motivation

Although stress has been a longstanding issue in organizations and management studies, it has never been studied in relation to Public Service Motivation. This article therefore aims to integrate PSM into the job demands-job resources model of stress in order to determine whether PSM might contribute to stress in public organizations. Drawing upon original data from a questionnaire in a Swiss municipality, this study unsurprisingly shows that "red tape" is an antecedent of stress perception, whereas satisfaction with organizational support, positive feedback, and recognition significantly decrease the level of perceived stress. Astonishingly, the empirical results show that PSM is positively and significantly related to stress perception. By increasing individuals' expectations towards their jobs, PSM might thus contribute to increased pressure on public agents. Ultimately, this article investigates the "dark side" of PSM, which has been neglected by the literature thus far.

Do resources, justice administration practices and federalism have an impact on registered and sentenced crime prevalence ?

This contribution, based on a statistical approach, undertakes to link data on resources (personnel and financial means) and the working of the administration of penal justice (prosecution, sentencing) taking into account the nationality of those prosecuted. In order to be able to distinguish prosecution and sentencing practices of judicial authorities and possible processes of discrimination, diverse sources have been used such as data from court administrations, public finances and police forces, collected by the Swiss Federal Statistical Office and the Swiss Federal administration of finances. The authors discuss discrimination in prosecution and sentencing between Swiss residents and foreigners taking into account localization and resources regarding personnel and public finances.

Comparison of HIV-infected patients' characteristics, healthcare resources use and cost between native and migrant patients

OBJECTIVES: To assess whether patients' characteristics and healthcare resources consumption and costs were different between native and migrant populations in Switzerland. METHODS: All adult patients followed-up in the Swiss HIV-cohort study in our institution during 2000-2003 were considered. Patients' characteristics were retrieved from the cohort database. Hospital and outpatient resource use were extracted from individual charts and valued with 2002 tariffs. RESULTS: The 66 migrants were younger (29 +/- 8 years versus 37 +/- 11, p < 0.001), less often of male gender (38 % versus 70 %, p < 0.001), predominantly infected via heterosexual contact (87 % versus 52 %, p < 0.01), with lower mean CD4 level at enrollment (326 +/- 235 versus 437 +/- 305, p = 0.002) than their 200 native counterparts. Migrants had fewer hospitalizations, more frequent outpatient visits, laboratory tests, and lower total cost of care per year of follow-up (<euro> 2'215 +/- 4'206 versus 4'155 +/- 12'304, p = 0.037). Resource use and costs were significantly higher in people with < 200 CD4 cell counts in both groups. CONCLUSIONS: Migrant population had more advanced disease, more outpatient visits but less hospitalizations, resulting in lower costs of care when compared with native population.

Database resources for the tuberculosis community.

Access to online repositories for genomic and associated "-omics" datasets is now an essential part of everyday research activity. It is important therefore that the Tuberculosis community is aware of the databases and tools available to them online, as well as for the database hosts to know what the needs of the research community are. One of the goals of the Tuberculosis Annotation Jamboree, held in Washington DC on March 7th-8th 2012, was therefore to provide an overview of the current status of three key Tuberculosis resources, TubercuList (tuberculist.epfl.ch), TB Database (www.tbdb.org), and Pathosystems Resource Integration Center (PATRIC, www.patricbrc.org). Here we summarize some key updates and upcoming features in TubercuList, and provide an overview of the PATRIC site and its online tools for pathogen RNA-Seq analysis.

Permanent Genetic Resources added to Molecular Ecology Resources Database 1 August 2010-30 September 2010

This article documents the addition of 229 microsatellite marker loci to the Molecular Ecology Resources Database. Loci were developed for the following species: Acacia auriculiformis x Acacia mangium hybrid, Alabama argillacea, Anoplopoma fimbria, Aplochiton zebra, Brevicoryne brassicae, Bruguiera gymnorhiza, Bucorvus leadbeateri, Delphacodes detecta, Tumidagena minuta, Dictyostelium giganteum, Echinogammarus berilloni, Epimedium sagittatum, Fraxinus excelsior, Labeo chrysophekadion, Oncorhynchus clarki lewisi, Paratrechina longicornis, Phaeocystis antarctica, Pinus roxburghii and Potamilus capax. These loci were cross-tested on the following species: Acacia peregrinalis, Acacia crassicarpa, Bruguiera cylindrica, Delphacodes detecta, Tumidagena minuta, Dictyostelium macrocephalum, Dictyostelium discoideum, Dictyostelium purpureum, Dictyostelium mucoroides, Dictyostelium rosarium, Polysphondylium pallidum, Epimedium brevicornum, Epimedium koreanum, Epimedium pubescens, Epimedium wushanese and Fraxinus angustifolia.

What makes a host profitable? Parasites balance host nutritive resources against immunity.

Numerous host qualities can modulate parasite fitness, and among these, host nutritive resources and immunity are of prime importance. Indeed, parasite fitness increases with the amount of nutritive resources extracted from the host body and decreases with host immune response. To maximize fitness, parasites have therefore to balance these two host components. Yet, because host nutritive resources and immunity both increase with host body condition, it is unclear whether parasites perform better on hosts in prime, intermediate, or poor condition. We investigated blood meal size and survival of the ectoparasitic louse fly Crataerina melbae in relation to body condition and cutaneous immune response of their Alpine swift (Apus melba) nestling hosts. Louse flies took a smaller blood meal and lived a shorter period of time when feeding on nestlings that were experimentally food deprived or had their cutaneous immune response boosted with methionine. Consistent with these results, louse fly survival was the highest when feeding on nonexperimental nestlings in intermediate body condition. Our findings emphasize that although hosts in poor condition had a reduced immunocompetence, parasites may have avoided them because individuals in poor condition did not provide adequate resources. These findings highlight the fact that giving host immunocompetence primary consideration can result in a biased appraisal of host-parasite interactions.