Bayesian models for finding and grouping junctions

In this paper, we propose two Bayesian methods for detecting and grouping junctions. Our junction detection method evolves from the Kona approach, and it is based on a competitive greedy procedure inspired in the region competition method. Then, junction grouping is accomplished by finding connecting paths between pairs of junctions. Path searching is performed by applying a Bayesian A* algorithm that has been recently proposed. Both methods are efficient and robust, and they are tested with synthetic and real images.

Cancer mortality inequalities in urban areas: a Bayesian small area analysis in Spanish cities

Background: Intra-urban inequalities in mortality have been infrequently analysed in European contexts. The aim of the present study was to analyse patterns of cancer mortality and their relationship with socioeconomic deprivation in small areas in 11 Spanish cities. Methods: It is a cross-sectional ecological design using mortality data (years 1996-2003). Units of analysis were the census tracts. A deprivation index was calculated for each census tract. In order to control the variability in estimating the risk of dying we used Bayesian models. We present the RR of the census tract with the highest deprivation vs. the census tract with the lowest deprivation. Results: In the case of men, socioeconomic inequalities are observed in total cancer mortality in all cities, except in Castellon, Cordoba and Vigo, while Barcelona (RR = 1.53 95%CI 1.42-1.67), Madrid (RR = 1.57 95%CI 1.49-1.65) and Seville (RR = 1.53 95%CI 1.36-1.74) present the greatest inequalities. In general Barcelona and Madrid, present inequalities for most types of cancer. Among women for total cancer mortality, inequalities have only been found in Barcelona and Zaragoza. The excess number of cancer deaths due to socioeconomic deprivation was 16,413 for men and 1,142 for women. Conclusion: This study has analysed inequalities in cancer mortality in small areas of cities in Spain, not only relating this mortality with socioeconomic deprivation, but also calculating the excess mortality which may be attributed to such deprivation. This knowledge is particularly useful to determine which geographical areas in each city need intersectorial policies in order to promote a healthy environment.

Beta diversity at multiple hierarchical levels: explaining the high diversity of scarab beetles in tropical montane forests

Aim: High gamma diversity in tropical montane forests may be ascribed to high geographical turnover of community composition, resulting from population isolation that leads to speciation. We studied the evolutionary processes responsible for diversity and turnover in assemblages of tropical scarab beetles (Scarabaeidae) by assessing DNA sequence variation at multiple hierarchical levels. Location: A 300-km transect across six montane forests (900–1100 m) in Costa Rica. Methods: Assemblages of Scarabaeidae (subfamilies Dynastinae, Rutelinae, Melolonthinae) including 118 morphospecies and > 500 individuals were sequenced for the cox1 gene to establish species limits with a mixed Yule–coalescent method. A species-level phylogenetic tree was constructed from cox1 and rrnL genes. Total diversity and turnover among assemblages were then assessed at three hierarchical levels: haplotypes, species and higher clades. Results: DNA-based analyses showed high turnover among communities at all hierarchical levels. Turnover was highest at the haplotype level (community similarity 0.02–0.12) and decreased with each step of the hierarchy (species: 0.21–0.46; clades: 0.41–0.43). Both compositional and phylogenetic similarities of communities were geographically structured, but turnover was not correlated with distance among forests. When three major clades were investigated separately, communities of Dynastinae showed consistently higher alpha diversity, larger species ranges and lower turnover than Rutelinae and Melolonthinae. Main conclusions: Scarab communities of montane forests show evidence of evolutionary persistence of communities in relative isolation, presumably tracking suitable habitats elevationally to accommodate climatic changes. Patterns of diversity on all hierarchical levels seem to be determined by restricted dispersal, and differences in Dynastinae could be explained by their greater dispersal ability. Community-wide DNA sequencing across multiple lineages and hierarchical levels reveals the evolutionary processes that led to high beta diversity in tropical montane forests through time.

Feature selection by multi-objective optimisation: Application to network anomaly detection by hierarchical self-organising maps

Feature selection is an important and active issue in clustering and classification problems. By choosing an adequate feature subset, a dataset dimensionality reduction is allowed, thus contributing to decreasing the classification computational complexity, and to improving the classifier performance by avoiding redundant or irrelevant features. Although feature selection can be formally defined as an optimisation problem with only one objective, that is, the classification accuracy obtained by using the selected feature subset, in recent years, some multi-objective approaches to this problem have been proposed. These either select features that not only improve the classification accuracy, but also the generalisation capability in case of supervised classifiers, or counterbalance the bias toward lower or higher numbers of features that present some methods used to validate the clustering/classification in case of unsupervised classifiers. The main contribution of this paper is a multi-objective approach for feature selection and its application to an unsupervised clustering procedure based on Growing Hierarchical Self-Organising Maps (GHSOMs) that includes a new method for unit labelling and efficient determination of the winning unit. In the network anomaly detection problem here considered, this multi-objective approach makes it possible not only to differentiate between normal and anomalous traffic but also among different anomalies. The efficiency of our proposals has been evaluated by using the well-known DARPA/NSL-KDD datasets that contain extracted features and labelled attacks from around 2 million connections. The selected feature sets computed in our experiments provide detection rates up to 99.8% with normal traffic and up to 99.6% with anomalous traffic, as well as accuracy values up to 99.12%.