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.

Urban image classification with semisupervised multiscale cluster kernels

This paper presents a semisupervised support vector machine (SVM) that integrates the information of both labeled and unlabeled pixels efficiently. Method's performance is illustrated in the relevant problem of very high resolution image classification of urban areas. The SVM is trained with the linear combination of two kernels: a base kernel working only with labeled examples is deformed by a likelihood kernel encoding similarities between labeled and unlabeled examples. Results obtained on very high resolution (VHR) multispectral and hyperspectral images show the relevance of the method in the context of urban image classification. Also, its simplicity and the few parameters involved make the method versatile and workable by unexperienced users.

Automatic Mallampati Classification Using Active Appearance Models

Difficult tracheal intubation assessment is an important research topic in anesthesia as failed intubations are important causes of mortality in anesthetic practice. The modified Mallampati score is widely used, alone or in conjunction with other criteria, to predict the difficulty of intubation. This work presents an automatic method to assess the modified Mallampati score from an image of a patient with the mouth wide open. For this purpose we propose an active appearance models (AAM) based method and use linear support vector machines (SVM) to select a subset of relevant features obtained using the AAM. This feature selection step proves to be essential as it improves drastically the performance of classification, which is obtained using SVM with RBF kernel and majority voting. We test our method on images of 100 patients undergoing elective surgery and achieve 97.9% accuracy in the leave-one-out crossvalidation test and provide a key element to an automatic difficult intubation assessment system.

Parents living with HIV in a high-income country: do patients need specific support?

The number of HIV-infected persons with children and caregiving duties is likely to increase. From this statement, the present study was designed to establish how HIV infected caregivers organise their parenting routines and to determine their support needs. A further aim was to ascertain caregivers' perception of conspicuous behaviours displayed by their children. Finally, it sought to determine the extent to which the caregivers' assessment of their parenting activity is influenced by the required support and their children's perceived conspicuous behaviours. The study design was observational and cross-sectional. Sampling was based on the 7 HIV Outpatient Clinics associated with the national population-based Swiss HIV Cohort Study. It focused on persons living with HIV who are responsible for raising children below the age of 18. A total of 520 caregivers were approached and 261 participated. An anonymous, standardised, self-administered questionnaire was used for data collection. The data were analysed using descriptive statistical procedures and backward elimination multiple regression analysis. The 261 respondents cared for 406 children and adolescents under 18 years of age; the median age was 10 years. The caregivers' material resources were low. 70% had a net family income in a range below the median of Swiss net family income and 30% were dependent on welfare assistance. 73% were undergoing treatment with 86% reporting no physical impairments. The proportion of single caregivers was 34%. 92% of the children were living with their HIV infected caregivers. 80% of the children attended an institution such as a school or kindergarten during the day. 89% of the caregivers had access to social networks providing support. Nevertheless, caregivers required additional support in performing their parenting duties and indicated a need for assistance on the material level, in connection with legal problems and with participation in the labour market. 46% of the caregivers had observed one or more conspicuous behaviours displayed by their children, which indicates a challenging situation. However, most of these caregivers assessed their parenting activity very favourably. Backward elimination multiple regression analysis indicated that a smaller number of support needs, younger age of the eldest child and fewer physical impairments on the part of the caregiver enhance the caregivers' assessment of their parenting activity. Physicians should speak to caregivers living with HIV about their parenting responsibilities and provide the necessary scope for this subject in their consultation sessions. Physicians are in a position to draw their patients' attention to the services available to them.

Identification and ethnic diversity underlie support for multicultural rights : a multilevel analysis in Bulgaria

Bulgaria is historically a multicultural society, composed of the Bulgarian (ethnic) majority and a number of ethnic minorities among which Bulgarian Turks and Roma are the largest. Both minority communities are stigmatized in contemporary Bulgaria, though to different degrees and for different reasons. Ethnic minorities' rights to preserve their culture, customs, and language are a topic of contentious debate. The purpose of this study was to examine individual- and context-level antecedents of the ethnic Bulgarian majority's support for multicultural rights of ethnic minorities. Multilevel regression analyses were conducted with International Social Survey Programme ISSP 2003 data (N = 920 in 28 Bulgarian districts). At the individual-level, an ethnic conception of the nation and anti-Roma symbolic prejudice were negatively related to support for multicultural rights, whereas national identification was positively related to the support of these rights. Over and above individual-level effects, and in line with recent extensions of intergroup contact theory, thepercentage ofBulgarianTurks withindistricts was positively related to support for multicultural rights. Importantly, support for multicultural rights was particularly high in districts characterized by ethnic diversity, that is, in districts with high proportions of both Bulgarian Turks and Roma. The beneficial effects of ethnic diversity and theoretical implications of findings are discussed.