Search

The selection of attributes becomes more important, but also more difficult, as the size and dimensionality of data sets grows, particularly in bioinformatics. Targeted Projection Pursuit is a dimension reduction technique previously applied to visualising high-dimensional data; here it is applied to the problem of feature selection. The technique avoids searching the powerset of possible feature combinations by using perceptron learning and attraction-repulsion algorithms to find projections that separate classes in the data. The technique is tested on a range of gene expression data sets. It is found that the classification generalisation performance of the features selected by TPP compares well with standard wrapper and filter approaches, the selection of features generalises more robustly than either, and its time efficiency scales to larger numbers of attributes better than standard searches.

The maximum data rate that can be achieved in diffuse indoor optical wireless communication (OWC) is limited due to the effect of intersymbol interference (ISI). The adverse effect of ISI on the system performance can be minimised using a channel equaliser at the receiver. In this study, digital signal detection is formulated as a classification problem and hence a classifier is adopted at the receiver. The bit error performance of classifiers with non-linear decision boundary including a multilayer perceptron (MLP), a support vector machine (SVM), the radial basis function (RBF), and the Bayesian classifier is studied along with traditional equaliser and reported here. The MLP offers the best performance; however there is trade-off between the performance and complexity especially at highly diffuse channel.

This paper describes the conversion of partial hospitals into recovery-oriented programs as part of system transformation. Steps included: participatory planning with stakeholders; strength based assessment of resources and needs; technical assistance; and changing funding strategies. Over a period of 8 years, use of partial hospitals decreased as persons with serious mental illnesses were transitioned to community integrated recovery centers. Preliminary outcomes suggest that these programs are more effective in engaging people in the community activities of their choice, confirming previous findings that showed that partial hospitals can be converted to recovery-oriented programs that focus more directly on promoting community inclusion.

Systems biology and bioinformatics are now major fields for productive research. DNA microarrays and other array technologies and genome sequencing have advanced to the point that it is now possible to monitor gene expression on a genomic scale. Gene expression analysis is discussed and some important clustering techniques are considered. The patterns identified in the data suggest similarities in the gene behavior, which provides useful information for the gene functionalities. We discuss measures for investigating the homogeneity of gene expression data in order to optimize the clustering process. We contribute to the knowledge of functional roles and regulation of E. coli genes by proposing a classification of these genes based on consistently correlated genes in expression data and similarities of gene expression patterns. A new visualization tool for targeted projection pursuit and dimensionality reduction of gene expression data is demonstrated.

Many algorithms for graph layout have been devised over the last 30 years spanning both the graph drawing and information visualisation communities. This article first reviews the advances made in the field of graph drawing that have then often been applied by the information visualisation community. There then follows a discussion of a range of techniques developed specifically for graph visualisations. Graph drawing algorithms are categorised into the following approaches: force-directed layouts, the use of dimension reduction in graph layout and computational improvements including multi-level techniques. Methods developed specifically for graph visualisation often make use of node-attributes and are categorised based on whether the attributes are used to introduce constraints to the layout, provide a clustered view or define an explicit representation in two-dimensional space. The similarities and distinctions between these techniques are examined and the aim is to provide a detailed assessment of currently available graph layout techniques, specifically how they can be used by visualisation practitioners, and to motivate further research in the area.

Summary A tool is introduced that uses a novel technique to enable users to explore two-dimensional views of high dimensional gene expression data sets. Unlike other such tools, the interface is intuitive and efficient, allowing the user to easily select views that meet their requirements. The tool is tested on publicly available gene expression data sets and demonstrated to find views that show the seperation of gene expression data sets into classes more effectively than standard dimension-reduction methods.