Multitask learning of environmental spatial data

The present research deals with an application of artificial neural networks for multitask learning from spatial environmental data. The real case study (sediments contamination of Geneva Lake) consists of 8 pollutants. There are different relationships between these variables, from linear correlations to strong nonlinear dependencies. The main idea is to construct a subsets of pollutants which can be efficiently modeled together within the multitask framework. The proposed two-step approach is based on: 1) the criterion of nonlinear predictability of each variable ?k? by analyzing all possible models composed from the rest of the variables by using a General Regression Neural Network (GRNN) as a model; 2) a multitask learning of the best model using multilayer perceptron and spatial predictions. The results of the study are analyzed using both machine learning and geostatistical tools.

Multiple Geotechnological Tools Applied to Digital Mapping of Tropical Soils

ABSTRACT In recent years, geotechnologies as remote and proximal sensing and attributes derived from digital terrain elevation models indicated to be very useful for the description of soil variability. However, these information sources are rarely used together. Therefore, a methodology for assessing and specialize soil classes using the information obtained from remote/proximal sensing, GIS and technical knowledge has been applied and evaluated. Two areas of study, in the State of São Paulo, Brazil, totaling approximately 28.000 ha were used for this work. First, in an area (area 1), conventional pedological mapping was done and from the soil classes found patterns were obtained with the following information: a) spectral information (forms of features and absorption intensity of spectral curves with 350 wavelengths -2,500 nm) of soil samples collected at specific points in the area (according to each soil type); b) obtaining equations for determining chemical and physical properties of the soil from the relationship between the results obtained in the laboratory by the conventional method, the levels of chemical and physical attributes with the spectral data; c) supervised classification of Landsat TM 5 images, in order to detect changes in the size of the soil particles (soil texture); d) relationship between classes relief soils and attributes. Subsequently, the obtained patterns were applied in area 2 obtain pedological classification of soils, but in GIS (ArcGIS). Finally, we developed a conventional pedological mapping in area 2 to which was compared with a digital map, ie the one obtained only with pre certain standards. The proposed methodology had a 79 % accuracy in the first categorical level of Soil Classification System, 60 % accuracy in the second category level and became less useful in the categorical level 3 (37 % accuracy).

Postmortem dynamic CT-angiography : P30

Purpose: Forensic imaging and especially forensic radiology is a new trend in forensic medicine. More and more forensic institutes set up their own CT-scanner in order to perform postmortem cross-sectional imaging. Due to this trend, a new subspecialty was born: the forensic radiology. To image the vascular system after death, a postmortem CT- angiography can be performed. Methods and materials: In the Institute of Forensic Medicine in Lausanne, a science group has been created with specialists of different medical fields that has set up a new technique of forensic CT-angiography. The method consists in the creation of a postmortem circulation by the use of a modified heart lung machine. As circulating liquid Angiofil, an oily contrast agent, is injected. Results: With the aid of this technique, the whole vascular system of a deceased person can be imaged in detail without autopsy. The circulating contrast allows demonstrating the vascular system when it is under pressure, similarly to living patients. First experiences showed, that vascular pathologies such as cardiac tamponade and aortic dissection can be well demonstrated. Since the oily Angiofil strictly remains in the intravascular space, no artifacts had been observed during the CT-examination and the later performed autopsy. Conclusion: Post-mortem dynamic CT angiography is of great advantage in forensic pathology, because the detailed mapping of the entire vascular system is almost impossible with conventional autopsy tools.

GIS - Based Decision and Outreach Tools for Aggregate Source Management, September 2008

This research project combined various datasets, existing and created for this project, into an Interactive Mapping Service (IMS) for use by Iowa DOT personnel, county planning and zoning departments and the public in order to make more informed decisions regarding aggregate sources and future access to them. Iowa DOT Technical Advisory Committee meetings were held, along with public forum presentations, in order to understand better the social, ecological and economic limitations to extracting aggregate. The information needed by potential users was conveyed and integrated into a single informational source, the Aggregate Planning IMS.

European guidelines for the development of mechanisms/tools supporting the individual demand for training through use of training Vouchers

El projecte de TAV es basa en la transferència d'un sistema de vals de capacitació, i en com aquest sistema de vals és adaptable a altres països o regions. En el document s'analitzen diferents conceptes teòrics sobre la conveniència o no de la implantació d'aquest sistema. A més a més, aquest document és una guia per a les organitzacions interessades en l'adaptació del sistema de vals formatius al seu territori, mostrant els passos a seguir i oferint eines útils per aconseguir-ho.

E-tools for civic education : experiences from the election campaign in Switzerland 2007

Political participation is often very low in Switzerland especially among students and young citizens. In the run-up to the Swiss parliamentary election in October 2007 several online tools and campaigns were developed with the aim to increase not only the level of information about the political programs of parties and candidates, but also the electoral participation of younger citizens. From a practical point of view this paper will describe the development, marketing efforts and the distribution as well as the use of two of these tools : the so-called "Parteienkompass" (party compass) and the "myVote"-tool - an online voting assistance tool based on an issue-matching system comparing policy preferences between voters and candidates on an individual level. We also havea look at similar tools stemming from Voting Advice Applications (VAA) in other countries in Western Europe. The paper closes with the results of an evaluation and an outlook to further developments and on-going projects in the near future in Switzerland.

Benchmarking therapeutic drug monitoring software: A systematic evaluation of available computer tools

Introduction: Therapeutic drug monitoring (TDM) aims at optimizing treatment by individualizing dosage regimen based on measurement of blood concentrations. Maintaining concentrations within a target range requires pharmacokinetic and clinical capabilities. Bayesian calculation represents a gold standard in TDM approach but requires computing assistance. In the last decades computer programs have been developed to assist clinicians in this assignment. The aim of this benchmarking was to assess and compare computer tools designed to support TDM clinical activities.¦Method: Literature and Internet search was performed to identify software. All programs were tested on common personal computer. Each program was scored against a standardized grid covering pharmacokinetic relevance, user-friendliness, computing aspects, interfacing, and storage. A weighting factor was applied to each criterion of the grid to consider its relative importance. To assess the robustness of the software, six representative clinical vignettes were also processed through all of them.¦Results: 12 software tools were identified, tested and ranked. It represents a comprehensive review of the available software's characteristics. Numbers of drugs handled vary widely and 8 programs offer the ability to the user to add its own drug model. 10 computer programs are able to compute Bayesian dosage adaptation based on a blood concentration (a posteriori adjustment) while 9 are also able to suggest a priori dosage regimen (prior to any blood concentration measurement), based on individual patient covariates, such as age, gender, weight. Among those applying Bayesian analysis, one uses the non-parametric approach. The top 2 software emerging from this benchmark are MwPharm and TCIWorks. Other programs evaluated have also a good potential but are less sophisticated (e.g. in terms of storage or report generation) or less user-friendly.¦Conclusion: Whereas 2 integrated programs are at the top of the ranked listed, such complex tools would possibly not fit all institutions, and each software tool must be regarded with respect to individual needs of hospitals or clinicians. Interest in computing tool to support therapeutic monitoring is still growing. Although developers put efforts into it the last years, there is still room for improvement, especially in terms of institutional information system interfacing, user-friendliness, capacity of data storage and report generation.

Benchmarking therapeutic drug monitoring software: A systematic evaluation of available computer tools

Objectives: Therapeutic drug monitoring (TDM) aims at optimizing treatment by individualizing dosage regimen based on blood concentrations measurement. Maintaining concentrations within a target range requires pharmacokinetic (PK) and clinical capabilities. Bayesian calculation represents a gold standard in TDM approach but requires computing assistance. The aim of this benchmarking was to assess and compare computer tools designed to support TDM clinical activities.¦Methods: Literature and Internet were searched to identify software. Each program was scored against a standardized grid covering pharmacokinetic relevance, user-friendliness, computing aspects, interfacing, and storage. A weighting factor was applied to each criterion of the grid to consider its relative importance. To assess the robustness of the software, six representative clinical vignettes were also processed through all of them.¦Results: 12 software tools were identified, tested and ranked. It represents a comprehensive review of the available software characteristics. Numbers of drugs handled vary from 2 to more than 180, and integration of different population types is available for some programs. Nevertheless, 8 programs offer the ability to add new drug models based on population PK data. 10 computer tools incorporate Bayesian computation to predict dosage regimen (individual parameters are calculated based on population PK models). All of them are able to compute Bayesian a posteriori dosage adaptation based on a blood concentration while 9 are also able to suggest a priori dosage regimen, only based on individual patient covariates. Among those applying Bayesian analysis, MM-USC*PACK uses a non-parametric approach. The top 2 programs emerging from this benchmark are MwPharm and TCIWorks. Others programs evaluated have also a good potential but are less sophisticated or less user-friendly.¦Conclusions: Whereas 2 software packages are ranked at the top of the list, such complex tools would possibly not fit all institutions, and each program must be regarded with respect to individual needs of hospitals or clinicians. Programs should be easy and fast for routine activities, including for non-experienced users. Although interest in TDM tools is growing and efforts were put into it in the last years, there is still room for improvement, especially in terms of institutional information system interfacing, user-friendliness, capability of data storage and automated report generation.

Machine learning analysis and modeling of interest rate curves

The present research deals with the review of the analysis and modeling of Swiss franc interest rate curves (IRC) by using unsupervised (SOM, Gaussian Mixtures) and supervised machine (MLP) learning algorithms. IRC are considered as objects embedded into different feature spaces: maturities; maturity-date, parameters of Nelson-Siegel model (NSM). Analysis of NSM parameters and their temporal and clustering structures helps to understand the relevance of model and its potential use for the forecasting. Mapping of IRC in a maturity-date feature space is presented and analyzed for the visualization and forecasting purposes.

Diagnostic Tools for Identifying Sleepy Drivers in the Field, RB21-012, 2013

The overarching goal of this project was to identify and evaluate cognitive and behavioral indices that are sensitive to sleep deprivation and may help identify commercial motor vehicle drivers (CMV) who are at-risk for driving in a sleep deprived state and may prove useful in field tests administered by officers. To that end, we evaluated indices of driver physiognomy (e.g., yawning, droopy eyelids, etc.) and driver behavioral/cognitive state (e.g. distracted driving) and the sensitivity of these indices to objective measures of sleep deprivation. The measures of sleep deprivation were sampled on repeated occasions over a period of 3.5-months in each of 44 drivers diagnosed with Obstructive Sleep Apnea (OSA) and 22 controls (matched for gender, age within 5 years, education within 2 years, and county of residence for rural vs. urban driving). Comprehensive analyses showed that specific dimensions of driver physiognomy associated with sleepiness in previous research and face-valid composite scores of sleepiness did not: 1) distinguish participants with OSA from matched controls; 2) distinguish participants before and after PAP treatment including those who were compliant with their treatment; 3) predict levels of sleep deprivation acquired objectively from actigraphy watches, not even among those chronically sleep deprived. Those findings are consistent with large individual differences in driver physiognomy. In other words, when individuals were sleep deprived as confirmed by actigraphy watch output they did not show consistently reliable behavioral markers of being sleep deprived. This finding held whether each driver was compared to him/herself with adequate and inadequate sleep, and even among chronically sleep deprived drivers. The scientific evidence from this research study does not support the use of driver physiognomy as a valid measure of sleep deprivation or as a basis to judge whether a CMV driver is too fatigued to drive, as on the current Fatigued Driving Evaluation Checklist.. Fair and accurate determinations of CMV driver sleepiness in the field will likely require further research on alternative strategies that make use of a combination of information sources besides driver physiognomy, including work logs, actigraphy, in vehicle data recordings, GPS data on vehicle use, and performance tests.

Advanced geostatistical and machine-learning models for spatial data analysis of radioactively contaminated regions

Radioactive soil-contamination mapping and risk assessment is a vital issue for decision makers. Traditional approaches for mapping the spatial concentration of radionuclides employ various regression-based models, which usually provide a single-value prediction realization accompanied (in some cases) by estimation error. Such approaches do not provide the capability for rigorous uncertainty quantification or probabilistic mapping. Machine learning is a recent and fast-developing approach based on learning patterns and information from data. Artificial neural networks for prediction mapping have been especially powerful in combination with spatial statistics. A data-driven approach provides the opportunity to integrate additional relevant information about spatial phenomena into a prediction model for more accurate spatial estimates and associated uncertainty. Machine-learning algorithms can also be used for a wider spectrum of problems than before: classification, probability density estimation, and so forth. Stochastic simulations are used to model spatial variability and uncertainty. Unlike regression models, they provide multiple realizations of a particular spatial pattern that allow uncertainty and risk quantification. This paper reviews the most recent methods of spatial data analysis, prediction, and risk mapping, based on machine learning and stochastic simulations in comparison with more traditional regression models. The radioactive fallout from the Chernobyl Nuclear Power Plant accident is used to illustrate the application of the models for prediction and classification problems. This fallout is a unique case study that provides the challenging task of analyzing huge amounts of data ('hard' direct measurements, as well as supplementary information and expert estimates) and solving particular decision-oriented problems.

Applying machine learning methods to avalanche forecasting

Avalanche forecasting is a complex process involving the assimilation of multiple data sources to make predictions over varying spatial and temporal resolutions. Numerically assisted forecasting often uses nearest neighbour methods (NN), which are known to have limitations when dealing with high dimensional data. We apply Support Vector Machines to a dataset from Lochaber, Scotland to assess their applicability in avalanche forecasting. Support Vector Machines (SVMs) belong to a family of theoretically based techniques from machine learning and are designed to deal with high dimensional data. Initial experiments showed that SVMs gave results which were comparable with NN for categorical and probabilistic forecasts. Experiments utilising the ability of SVMs to deal with high dimensionality in producing a spatial forecast show promise, but require further work.