Multi-scale support vector algorithms for hot spot detection and modelling

The algorithmic approach to data modelling has developed rapidly these last years, in particular methods based on data mining and machine learning have been used in a growing number of applications. These methods follow a data-driven methodology, aiming at providing the best possible generalization and predictive abilities instead of concentrating on the properties of the data model. One of the most successful groups of such methods is known as Support Vector algorithms. Following the fruitful developments in applying Support Vector algorithms to spatial data, this paper introduces a new extension of the traditional support vector regression (SVR) algorithm. This extension allows for the simultaneous modelling of environmental data at several spatial scales. The joint influence of environmental processes presenting different patterns at different scales is here learned automatically from data, providing the optimum mixture of short and large-scale models. The method is adaptive to the spatial scale of the data. With this advantage, it can provide efficient means to model local anomalies that may typically arise in situations at an early phase of an environmental emergency. However, the proposed approach still requires some prior knowledge on the possible existence of such short-scale patterns. This is a possible limitation of the method for its implementation in early warning systems. The purpose of this paper is to present the multi-scale SVR model and to illustrate its use with an application to the mapping of Cs137 activity given the measurements taken in the region of Briansk following the Chernobyl accident.

Housewife or working mum--each to her own? The relevance of societal factors in the association between social roles and alcohol use among mothers in 16 industrialized countries.

Aims To investigate whether differences in gender-income equity at country level explain national differences in the links between alcohol use, and the combination of motherhood and paid labour. Design Cross-sectional data in 16 established market economies participating in the Gender, Alcohol and Culture: An International Study (GenACIS) study. Setting Population surveys. Participants A total of 12 454 mothers (aged 25-49 years). Measurements Alcohol use was assessed as the quantity per drinking day. Paid labour, having a partner, gender-income ratio at country level and the interaction between individual and country characteristics were regressed on alcohol consumed per drinking day using multi-level modelling. Findings Mothers with a partner who were in paid labour reported consuming more alcohol on drinking days than partnered housewives. In countries with high gender-income equity, mothers with a partner who were in paid labour drank less alcohol per occasion, while alcohol use was higher among working partnered mothers living in countries with lower income equity. Conclusion In countries which facilitate working mothers, daily alcohol use decreases as female social roles increase; in contrast, in countries where there are fewer incentives for mothers to remain in work, the protective effect of being a working mother (with partner) on alcohol use is weaker. These data suggest that a country's investment in measures to improve the compatibility of motherhood and paid labour may reduce women's alcohol use.

Analyse des risques d'extinction des populations naturelles

A review of extinction risk analysis and viability methods is presented. The importance of environmental, demographic and genetic uncertainties, as well as the role of catastrophes are successively considered, and different approaches aiming at the integration of these risk factors in predictive population dynamic models are discussed.