Uncertainty Quantification with Support Vector Regression Prediction Models

Uncertainty quantification of petroleum reservoir models is one of the present challenges, which is usually approached with a wide range of geostatistical tools linked with statistical optimisation or/and inference algorithms. The paper considers a data driven approach in modelling uncertainty in spatial predictions. Proposed semi-supervised Support Vector Regression (SVR) model has demonstrated its capability to represent realistic features and describe stochastic variability and non-uniqueness of spatial properties. It is able to capture and preserve key spatial dependencies such as connectivity, which is often difficult to achieve with two-point geostatistical models. Semi-supervised SVR is designed to integrate various kinds of conditioning data and learn dependences from them. A stochastic semi-supervised SVR model is integrated into a Bayesian framework to quantify uncertainty with multiple models fitted to dynamic observations. The developed approach is illustrated with a reservoir case study. The resulting probabilistic production forecasts are described by uncertainty envelopes.

Spatial variability of the chemical, physical and biological properties in lowland cultivated with irrigated rice

In the areas where irrigated rice is grown in the south of Brazil, few studies have been carried out to investigate the spatial variability structure of soil properties and to establish new forms of soil management as well as determine soil corrective and fertilizer applications. In this sense, this study had the objective of evaluating the spatial variability of chemical, physical and biological soil properties in a lowland area under irrigated rice cultivation in the conventional till system. For this purpose, a 10 x 10 m grid of 100 points was established, in an experimental field of the Embrapa Clima Temperado, in the County of Capão do Leão, State of Rio Grande do Sul. The spatial variability structure was evaluated by geostatistical tools and the number of subsamples required to represent each soil property in future studies was calculated using classical statistics. Results showed that the spatial variability structure of sand, silt, SMP index, cation exchange capacity (pH 7.0), Al3+ and total N properties could be detected by geostatistical analysis. A pure nugget effect was observed for the nutrients K, S and B, as well as macroporosity, mean weighted diameter of aggregates, and soil water storage. The cross validation procedure, based on linear regression and the determination coefficient, was more efficient to evaluate the quality of the adjusted mathematical model than the degree of spatial dependence. It was also concluded that the combination of classical with geostatistics can in many cases simplify the soil sampling process without losing information quality.

Pedotransfer functions related to spatial variability of water retention attributes for lowland soils

The estimation of non available soil variables through the knowledge of other related measured variables can be achieved through pedotransfer functions (PTF) mainly saving time and reducing cost. Great differences among soils, however, can yield non desirable results when applying this method. This study discusses the application of developed PTFs by several authors using a variety of soils of different characteristics, to evaluate soil water contents of two Brazilian lowland soils. Comparisons are made between PTF evaluated data and field measured data, using statistical and geostatistical tools, like mean error, root mean square error, semivariogram, cross-validation, and regression coefficient. The eight tested PTFs to evaluate gravimetric soil water contents (Ug) at the tensions of 33 kPa and 1,500 kPa presented a tendency to overestimate Ug 33 kPa and underestimate Ug1,500 kPa. The PTFs were ranked according to their performance and also with respect to their potential in describing the structure of the spatial variability of the set of measured values. Although none of the PTFs have changed the distribution pattern of the data, all resulted in mean and variance statistically different from those observed for all measured values. The PTFs that presented the best predictive values of Ug33 kPa and Ug1,500 kPa were not the same that had the best performance to reproduce the structure of spatial variability of these variables.

Geomodelling of a fluvial system with semi-supervised support vector regression

Fluvial deposits are a challenge for modelling flow in sub-surface reservoirs. Connectivity and continuity of permeable bodies have a major impact on fluid flow in porous media. Contemporary object-based and multipoint statistics methods face a problem of robust representation of connected structures. An alternative approach to model petrophysical properties is based on machine learning algorithm ? Support Vector Regression (SVR). Semi-supervised SVR is able to establish spatial connectivity taking into account the prior knowledge on natural similarities. SVR as a learning algorithm is robust to noise and captures dependencies from all available data. Semi-supervised SVR applied to a synthetic fluvial reservoir demonstrated robust results, which are well matched to the flow performance

Automatic Mapping and Classification of Spatial Environmental Data

The paper deals with the development and application of the generic methodology for automatic processing (mapping and classification) of environmental data. General Regression Neural Network (GRNN) is considered in detail and is proposed as an efficient tool to solve the problem of spatial data mapping (regression). The Probabilistic Neural Network (PNN) is considered as an automatic tool for spatial classifications. The automatic tuning of isotropic and anisotropic GRNN/PNN models using cross-validation procedure is presented. Results are compared with the k-Nearest-Neighbours (k-NN) interpolation algorithm using independent validation data set. Real case studies are based on decision-oriented mapping and classification of radioactively contaminated territories.

Retrospective feasibility study of simultaneous integrated boost in cervical cancer using Tomotherapy: the impact of organ motion and tumor regression.

BACKGROUND: Whole pelvis intensity modulated radiotherapy (IMRT) is increasingly being used to treat cervical cancer aiming to reduce side effects. Encouraged by this, some groups have proposed the use of simultaneous integrated boost (SIB) to target the tumor, either to get a higher tumoricidal effect or to replace brachytherapy. Nevertheless, physiological organ movement and rapid tumor regression throughout treatment might substantially reduce any benefit of this approach. PURPOSE: To evaluate the clinical target volume - simultaneous integrated boost (CTV-SIB) regression and motion during chemo-radiotherapy (CRT) for cervical cancer, and to monitor treatment progress dosimetrically and volumetrically to ensure treatment goals are met. METHODS AND MATERIALS: Ten patients treated with standard doses of CRT and brachytherapy were retrospectively re-planned using a helical Tomotherapy - SIB technique for the hypothetical scenario of this feasibility study. Target and organs at risk (OAR) were contoured on deformable fused planning-computed tomography and megavoltage computed tomography images. The CTV-SIB volume regression was determined. The center of mass (CM) was used to evaluate the degree of motion. The Dice's similarity coefficient (DSC) was used to assess the spatial overlap of CTV-SIBs between scans. A cumulative dose-volume histogram modeled estimated delivered doses. RESULTS: The CTV-SIB relative reduction was between 31 and 70%. The mean maximum CM change was 12.5, 9, and 3 mm in the superior-inferior, antero-posterior, and right-left dimensions, respectively. The CTV-SIB-DSC approached 1 in the first week of treatment, indicating almost perfect overlap. CTV-SIB-DSC regressed linearly during therapy, and by the end of treatment was 0.5, indicating 50% discordance. Two patients received less than 95% of the prescribed dose. Much higher doses to the OAR were observed. A multiple regression analysis showed a significant interaction between CTV-SIB reduction and OAR dose increase. CONCLUSIONS: The CTV-SIB had important regression and motion during CRT, receiving lower therapeutic doses than expected. The OAR had unpredictable shifts and received higher doses. The use of SIB without frequent adaptation of the treatment plan exposes cervical cancer patients to an unpredictable risk of under-dosing the target and/or overdosing adjacent critical structures. In that scenario, brachytherapy continues to be the gold standard approach.

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.

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.

Analysis and spatial modeling of the indoor radon Swiss data

The present research deals with an important public health threat, which is the pollution created by radon gas accumulation inside dwellings. The spatial modeling of indoor radon in Switzerland is particularly complex and challenging because of many influencing factors that should be taken into account. Indoor radon data analysis must be addressed from both a statistical and a spatial point of view. As a multivariate process, it was important at first to define the influence of each factor. In particular, it was important to define the influence of geology as being closely associated to indoor radon. This association was indeed observed for the Swiss data but not probed to be the sole determinant for the spatial modeling. The statistical analysis of data, both at univariate and multivariate level, was followed by an exploratory spatial analysis. Many tools proposed in the literature were tested and adapted, including fractality, declustering and moving windows methods. The use of Quan-tité Morisita Index (QMI) as a procedure to evaluate data clustering in function of the radon level was proposed. The existing methods of declustering were revised and applied in an attempt to approach the global histogram parameters. The exploratory phase comes along with the definition of multiple scales of interest for indoor radon mapping in Switzerland. The analysis was done with a top-to-down resolution approach, from regional to local lev¬els in order to find the appropriate scales for modeling. In this sense, data partition was optimized in order to cope with stationary conditions of geostatistical models. Common methods of spatial modeling such as Κ Nearest Neighbors (KNN), variography and General Regression Neural Networks (GRNN) were proposed as exploratory tools. In the following section, different spatial interpolation methods were applied for a par-ticular dataset. A bottom to top method complexity approach was adopted and the results were analyzed together in order to find common definitions of continuity and neighborhood parameters. Additionally, a data filter based on cross-validation was tested with the purpose of reducing noise at local scale (the CVMF). At the end of the chapter, a series of test for data consistency and methods robustness were performed. This lead to conclude about the importance of data splitting and the limitation of generalization methods for reproducing statistical distributions. The last section was dedicated to modeling methods with probabilistic interpretations. Data transformation and simulations thus allowed the use of multigaussian models and helped take the indoor radon pollution data uncertainty into consideration. The catego-rization transform was presented as a solution for extreme values modeling through clas-sification. Simulation scenarios were proposed, including an alternative proposal for the reproduction of the global histogram based on the sampling domain. The sequential Gaussian simulation (SGS) was presented as the method giving the most complete information, while classification performed in a more robust way. An error measure was defined in relation to the decision function for data classification hardening. Within the classification methods, probabilistic neural networks (PNN) show to be better adapted for modeling of high threshold categorization and for automation. Support vector machines (SVM) on the contrary performed well under balanced category conditions. In general, it was concluded that a particular prediction or estimation method is not better under all conditions of scale and neighborhood definitions. Simulations should be the basis, while other methods can provide complementary information to accomplish an efficient indoor radon decision making.

Ensemble methods for environmental data modelling with support vector regression

This paper investigates the use of ensemble of predictors in order to improve the performance of spatial prediction methods. Support vector regression (SVR), a popular method from the field of statistical machine learning, is used. Several instances of SVR are combined using different data sampling schemes (bagging and boosting). Bagging shows good performance, and proves to be more computationally efficient than training a single SVR model while reducing error. Boosting, however, does not improve results on this specific problem.

Maize yield and rainfall on different spatial and temporal scales in Southern Brazil

This study aimed to establish relationships between maize yield and rainfall on different temporal and spatial scales, in order to provide a basis for crop monitoring and modelling. A 16-year series of maize yield and daily rainfall from 11 municipalities and micro-regions of Rio Grande do Sul State was used. Correlation and regression analyses were used to determine associations between crop yield and rainfall for the entire crop cycle, from tasseling to 30 days after, and from 5 days before tasseling to 40 days after. Close relationships between maize yield and rainfall were found, particularly during the reproductive period (45-day period comprising the flowering and grain filling). Relationships were closer on a regional scale than at smaller scales. Implications of the crop-rainfall relationships for crop modelling are discussed.

Robust Nonlinear Mapping of Soil Contamination Using Support Regression

Spatial data analysis mapping and visualization is of great importance in various fields: environment, pollution, natural hazards and risks, epidemiology, spatial econometrics, etc. A basic task of spatial mapping is to make predictions based on some empirical data (measurements). A number of state-of-the-art methods can be used for the task: deterministic interpolations, methods of geostatistics: the family of kriging estimators (Deutsch and Journel, 1997), machine learning algorithms such as artificial neural networks (ANN) of different architectures, hybrid ANN-geostatistics models (Kanevski and Maignan, 2004; Kanevski et al., 1996), etc. All the methods mentioned above can be used for solving the problem of spatial data mapping. Environmental empirical data are always contaminated/corrupted by noise, and often with noise of unknown nature. That's one of the reasons why deterministic models can be inconsistent, since they treat the measurements as values of some unknown function that should be interpolated. Kriging estimators treat the measurements as the realization of some spatial randomn process. To obtain the estimation with kriging one has to model the spatial structure of the data: spatial correlation function or (semi-)variogram. This task can be complicated if there is not sufficient number of measurements and variogram is sensitive to outliers and extremes. ANN is a powerful tool, but it also suffers from the number of reasons. of a special type ? multiplayer perceptrons ? are often used as a detrending tool in hybrid (ANN+geostatistics) models (Kanevski and Maignank, 2004). Therefore, development and adaptation of the method that would be nonlinear and robust to noise in measurements, would deal with the small empirical datasets and which has solid mathematical background is of great importance. The present paper deals with such model, based on Statistical Learning Theory (SLT) - Support Vector Regression. SLT is a general mathematical framework devoted to the problem of estimation of the dependencies from empirical data (Hastie et al, 2004; Vapnik, 1998). SLT models for classification - Support Vector Machines - have shown good results on different machine learning tasks. The results of SVM classification of spatial data are also promising (Kanevski et al, 2002). The properties of SVM for regression - Support Vector Regression (SVR) are less studied. First results of the application of SVR for spatial mapping of physical quantities were obtained by the authorsin for mapping of medium porosity (Kanevski et al, 1999), and for mapping of radioactively contaminated territories (Kanevski and Canu, 2000). The present paper is devoted to further understanding of the properties of SVR model for spatial data analysis and mapping. Detailed description of the SVR theory can be found in (Cristianini and Shawe-Taylor, 2000; Smola, 1996) and basic equations for the nonlinear modeling are given in section 2. Section 3 discusses the application of SVR for spatial data mapping on the real case study - soil pollution by Cs137 radionuclide. Section 4 discusses the properties of the modelapplied to noised data or data with outliers.

Improved predictive mapping of indoor radon concentrations using ensemble regression trees based on automatic clustering of geological units.

PURPOSE: According to estimations around 230 people die as a result of radon exposure in Switzerland. This public health concern makes reliable indoor radon prediction and mapping methods necessary in order to improve risk communication to the public. The aim of this study was to develop an automated method to classify lithological units according to their radon characteristics and to develop mapping and predictive tools in order to improve local radon prediction. METHOD: About 240 000 indoor radon concentration (IRC) measurements in about 150 000 buildings were available for our analysis. The automated classification of lithological units was based on k-medoids clustering via pair-wise Kolmogorov distances between IRC distributions of lithological units. For IRC mapping and prediction we used random forests and Bayesian additive regression trees (BART). RESULTS: The automated classification groups lithological units well in terms of their IRC characteristics. Especially the IRC differences in metamorphic rocks like gneiss are well revealed by this method. The maps produced by random forests soundly represent the regional difference of IRCs in Switzerland and improve the spatial detail compared to existing approaches. We could explain 33% of the variations in IRC data with random forests. Additionally, the influence of a variable evaluated by random forests shows that building characteristics are less important predictors for IRCs than spatial/geological influences. BART could explain 29% of IRC variability and produced maps that indicate the prediction uncertainty. CONCLUSION: Ensemble regression trees are a powerful tool to model and understand the multidimensional influences on IRCs. Automatic clustering of lithological units complements this method by facilitating the interpretation of radon properties of rock types. This study provides an important element for radon risk communication. Future approaches should consider taking into account further variables like soil gas radon measurements as well as more detailed geological information.

Sources of variability of speech, spatial, and qualities of hearing scale (SSQ) scores in normal-hearing and hearing-impaired populations.

OBJECTIVE: To identify and quantify sources of variability in scores on the speech, spatial, and qualities of hearing scale (SSQ) and its short forms among normal-hearing and hearing-impaired subjects using a French-language version of the SSQ. DESIGN: Multi-regression analyses of SSQ scores were performed using age, gender, years of education, hearing loss, and hearing-loss asymmetry as predictors. Similar analyses were performed for each subscale (Speech, Spatial, and Qualities), for several SSQ short forms, and for differences in subscale scores. STUDY SAMPLE: One hundred normal-hearing subjects (NHS) and 230 hearing-impaired subjects (HIS). RESULTS: Hearing loss in the better ear and hearing-loss asymmetry were the two main predictors of scores on the overall SSQ, the three main subscales, and the SSQ short forms. The greatest difference between the NHS and HIS was observed for the Speech subscale, and the NHS showed scores well below the maximum of 10. An age effect was observed mostly on the Speech subscale items, and the number of years of education had a significant influence on several Spatial and Qualities subscale items. CONCLUSION: Strong similarities between SSQ scores obtained across different populations and languages, and between SSQ and short forms, underline their potential international use.

Environmental factors, spatial variation, and specific requirements of Chironomidae in Mediterranean reference streams

Chironomidae spatial distribution was investigated at 63 near-pristine sites in 22 catchments of the Iberian Mediterranean coast. We used partial redundancy analysis to study Chironomidae community responses to a number of environmental factors acting at several spatial scales. The percentage of variation explained by local factors (23.3%) was higher than that explained by geographical (8.5%) or regional factors(8%). Catchment area, longitude, pH, % siliceous rocks in the catchment, and altitude were the best predictors of Chironomidae assemblages. We used a k-means cluster analysis to classified sites into 3 major groups based on Chironomidae assemblages. These groups were explained mainly by longitudinal zonation and geographical position, and were defined as 1) siliceous headwater streams, 2) mid-altitude streams with small catchment areas, and 3) medium-sized calcareous streams. Distinct species assemblages with associated indicator taxa were established for each stream category using IndVal analysis. Species responses to previously identified key environmental variables were determined, and optima and tolerances were established by weighted average regression. Distinct ecological requirements were observed among genera and among species of the same genus. Some genera were restricted to headwater systems (e.g., Diamesa), whereas others (e.g., Eukiefferiella) had wider ecological preferences but with distinct distributions among congenerics. In the present period of climate change, optima and tolerances of species might be a useful tool to predict responses of different species to changes in significant environmental variables, such as temperature and hydrology.