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.

Missing value imputation in longitudinal measures of alcohol consumption

Attrition in longitudinal studies can lead to biased results. The study is motivated by the unexpected observation that alcohol consumption decreased despite increased availability, which may be due to sample attrition of heavy drinkers. Several imputation methods have been proposed, but rarely compared in longitudinal studies of alcohol consumption. The imputation of consumption level measurements is computationally particularly challenging due to alcohol consumption being a semi-continuous variable (dichotomous drinking status and continuous volume among drinkers), and the non-normality of data in the continuous part. Data come from a longitudinal study in Denmark with four waves (2003-2006) and 1771 individuals at baseline. Five techniques for missing data are compared: Last value carried forward (LVCF) was used as a single, and Hotdeck, Heckman modelling, multivariate imputation by chained equations (MICE), and a Bayesian approach as multiple imputation methods. Predictive mean matching was used to account for non-normality, where instead of imputing regression estimates, "real" observed values from similar cases are imputed. Methods were also compared by means of a simulated dataset. The simulation showed that the Bayesian approach yielded the most unbiased estimates for imputation. The finding of no increase in consumption levels despite a higher availability remained unaltered. Copyright (C) 2011 John Wiley & Sons, Ltd.

Approaches to identifying and quantifying polycyclic aromatic hydrocarbons of molecular weight 302 in diesel particulates

Among the PAH class of compounds, high molecular weight PAH are now considered as relevant cancer inducers, but not all of them have the same biological activity. However, their analysis is difficult, mainly due to the presence of numerous isomers and due to their low volatility. Retention indices (Ri) for 13 dibenzopyrenes and homologues were determined by high-resolution capillary gas chromatography (GC) with four different stationary phases: a 5% phenyl-substituted methylpolysiloxane column (DB-5 ms), a 35% phenyl-substituted methylpolysiloxane column (BPX-35), a 50% phenyl-substituted methylpolysiloxane column (BPX-50), and a 35% trifluoropropylmethyl polysiloxane stationary phase (Rtx-200). Correlations for retention on each phase were investigated by using 8 independent molecular descriptors. Ri has been shown to be linearly correlated to PAH volume, polarisability alpha, Hückel-pi energy on the four examined columns. Ionisation potential Ip is a fourth variable which improves the regression model for DB-5ms, BPX-35, and BPX-50 column. Correlation coefficients ranging from r2 = 0.935 to r2 = 0.952 are then observed. Application of these indices to the identification and quantification of PAH with MW 302 in certified diesel particulate matter SRM 1650a is presented and discussed. [Authors]

Association of urinary calcium excretion with serum calcium and vitamin D levels.

BACKGROUND AND OBJECTIVES: Population-based data on urinary calcium excretion are scarce. The association of serum calcium and circulating levels of vitamin D [25(OH)D2 or D3] with urinary calcium excretion in men and women from a population-based study was explored. DESIGN, SETTINGS, PARTICIPANTS, & MEASUREMENTS: Multivariable linear regression was used to explore factors associated with square root-transformed 24-hour urinary calcium excretion (milligrams per 24 hours) taken as the dependent variable with a focus on month-specific vitamin D tertiles and serum calcium in the Swiss Survey on Salt Study. RESULTS: In total, 624 men and 669 women were studied with mean ages of 49.2 and 47.0 years, respectively (age range=15-95 years). Mean urinary calcium excretion was higher in men than in women (183.05 versus 144.60 mg/24 h; P<0.001). In adjusted models, the association (95% confidence interval) of square root urinary calcium excretion with protein-corrected serum calcium was 1.78 (95% confidence interval, 1.21 to 2.34) mg/24 h per milligram per deciliter in women and 0.59 (95% confidence interval, -0.11 to 1.29) mg/24 h per milligram per deciliter in men. Men in the third 25(OH)D3 tertile had higher square root urinary calcium excretion than men in the first tertile (0.99; 95% confidence interval, 0.36 to 1.63 mg/24 h per nanogram per milliliter), and the corresponding association was 0.32 (95% confidence interval, -0.22 to 0.85) mg/24 h per nanogram per milliliter in women. These sex differences were more marked under conditions of high urinary sodium or urea excretions. CONCLUSIONS: There was a positive association of serum calcium with urinary calcium excretion in women but not men. Vitamin 25(OH)D3 was associated with urinary calcium excretion in men but not women. These results suggest important sex differences in the hormonal and dietary control of urinary calcium excretion.

General and Abdominal Adiposity and Risk of Death in Europe

Previous studies have relied predominantly on the body-mass index (BMI, the weight in kilograms divided by the square of the height in meters) to assess the association of adiposity with the risk of death, but few have examined whether the distribution of body fat contributes to the prediction of death. This study examined the association of BMI, waist circumference, and waist-to-hip ratio with the risk of death among 359,387 participants from nine countries in the European Prospective Investigation into Cancer and Nutrition (EPIC). A Cox regression analysis was used, with age as the time variable, and stratified the models according to study center and age at recruitment, with further adjustment for educational level, smoking status, alcohol consumption, physical activity, and height.

Melan-A-specific cytotoxic T cells are associated with tumor regression and autoimmunity following treatment with anti-CTLA-4.

PURPOSE: Ipilimumab is a monoclonal antibody that blocks the immune-inhibitory interaction between CTL antigen 4 (CTLA-4) and its ligands on T cells. Clinical trials in cancer patients with ipilimumab have shown promising antitumor activity, particularly in patients with advanced melanoma. Often, tumor regressions in these patients are correlated with immune-related side effects such as dermatitis, enterocolitis, and hypophysitis. Although these reactions are believed to be immune-mediated, the antigenic targets for the cellular or humoral immune response are not known. EXPERIMENTAL DESIGN: We enrolled patients with advanced melanoma in a phase II study with ipilimumab. One of these patients experienced a complete remission of his tumor. The specificity and functional properties of CD8-positive T cells in his peripheral blood, in regressing tumor tissue, and at the site of an immune-mediated skin rash were investigated. RESULTS: Regressing tumor tissue was infiltrated with CD8-positive T cells, a high proportion of which were specific for Melan-A. The skin rash was similarly infiltrated with Melan-A-specific CD8-positive T cells, and a dramatic (>30-fold) increase in Melan-A-specific CD8-positive T cells was apparent in peripheral blood. These cells had an effector phenotype and lysed Melan-A-expressing tumor cells. CONCLUSIONS: Our results show that Melan-A may be a major target for both the autoimmune and antitumor reactions in patients treated with anti-CTLA-4, and describe for the first time the antigen specificity of CD8-positive T cells that mediate tumor rejection in a patient undergoing treatment with an anti-CTLA-4 antibody. These findings may allow a better integration of ipilimumab into other forms of immunotherapy.

Serum uric acid and adiposity: deciphering causality using a bidirectional Mendelian randomization approach.

Although the relationship between serum uric acid (SUA) and adiposity is well established, the direction of the causality is still unclear in the presence of conflicting evidences. We used a bidirectional Mendelian randomization approach to explore the nature and direction of causality between SUA and adiposity in a population-based study of Caucasians aged 35 to 75 years. We used, as instrumental variables, rs6855911 within the SUA gene SLC2A9 in one direction, and combinations of SNPs within the adiposity genes FTO, MC4R and TMEM18 in the other direction. Adiposity markers included weight, body mass index, waist circumference and fat mass. We applied a two-stage least squares regression: a regression of SUA/adiposity markers on our instruments in the first stage and a regression of the response of interest on the fitted values from the first stage regression in the second stage. SUA explained by the SLC2A9 instrument was not associated to fat mass (regression coefficient [95% confidence interval]: 0.05 [-0.10, 0.19] for fat mass) contrasting with the ordinary least square estimate (0.37 [0.34, 0.40]). By contrast, fat mass explained by genetic variants of the FTO, MC4R and TMEM18 genes was positively and significantly associated to SUA (0.31 [0.01, 0.62]), similar to the ordinary least square estimate (0.27 [0.25, 0.29]). Results were similar for the other adiposity markers. Using a bidirectional Mendelian randomization approach in adult Caucasians, our findings suggest that elevated SUA is a consequence rather than a cause of adiposity.

Binding of d-methadone, l-methadone, and dl-methadone to proteins in plasma of healthy volunteers: role of the variants of alpha 1-acid glycoprotein.

The plasma concentrations of alpha 1-acid glycoprotein (AAG), albumin, triglycerides, cholesterol, and total proteins, as well as the plasma binding of racemic, d-methadone, and l-methadone were measured in 45 healthy subjects. The AAG phenotypes and the concentrations of AAG variants were also determined. The measured free fractions for racemic, d-methadone, and l-methadone were, respectively, 12.7% +/- 3.3%, 10.0% +/- 2.9%, and 14.2% +/- 3.2% (mean +/- SD). A significant correlation was obtained between the binding ratio (B/F) for dl-methadone and the total AAG concentration (r = 0.724; p less than 0.001). A multiple stepwise regression analysis showed that AAG was the main explanatory variable for the binding of the racemate. When concentrations of AAG variants were considered, a significant correlation was obtained between the binding ratio of dl-methadone and orosomucoid2 A concentration (r = 0.715; p less than 0.001), a weak correlation between dl-methadone and orosomucoid1 S concentration (r = 0.494; p less than 0.001), and no correlation between dl-methadone and orosomucoid1 F1 concentration (r = 0.049; not significant). Similar findings were obtained with the enantiomers. This study shows the importance of considering not only total AAG but also concentrations of AAG variants when measuring the binding of methadone and possibly of other drugs in plasma.

Analysis and estimative of schistosomiasis prevalence for the state of Minas Gerais, Brazil, using multiple regression with social and environmental spatial data

The aim of this work is to establish a relationship between schistosomiasis prevalence and social-environmental variables, in the state of Minas Gerais, Brazil, through multiple linear regression. The final regression model was established, after a variables selection phase, with a set of spatial variables which contains the summer minimum temperature, human development index, and vegetation type variables. Based on this model, a schistosomiasis risk map was built for Minas Gerais.

Adolescent substance-use assessment: methodological issues in the use of the ADAD (Adolescent Drug Abuse Diagnosis).

During the past twenty years, various instruments have been developed for the assessment of substance use in adolescents, mainly in the United States. However, few of them have been adapted to, and validated in, French-speaking populations. Consequently, although increasing alcohol and drug use among teenagers has become a major concern, the various health and social programs developed in response to this specific problem have received little attention with regard to follow-up and outcome assessment. A standardized multidimensional assessment instrument adapted for adolescents is needed to assess the individual needs of adolescents and assign them to the most appropriate treatment setting, to provide a single measurement within and across health and social systems, and to conduct treatment outcome evaluations. Moreover, having an available instrument makes it possible to develop longitudinal and trans-cultural research studies. For this reason, a French version of the Adolescent Drug Abuse Diagnosis (ADAD) was developed and validated at the University Child and Adolescent Psychiatric Clinic in Lausanne, Switzerland. This paper aims to discuss the methodological issues that we faced when using the ADAD instrument in a 4-year longitudinal study including adolescent substance users. Methodological aspects relating to the content and format of the instrument, the assessment administration and the statistical analyses are discussed.

Validation of Logistic Regression Models for Landslide Susceptibility Maps

A wide range of numerical models and tools have been developed over the last decades to support the decision making process in environmental applications, ranging from physical models to a variety of statistically-based methods. In this study, a landslide susceptibility map of a part of Three Gorges Reservoir region of China was produced, employing binary logistic regression analyses. The available information includes the digital elevation model of the region, geological map and different GIS layers including land cover data obtained from satellite imagery. The landslides were observed and documented during the field studies. The validation analysis is exploited to investigate the quality of mapping.