An unsupervised approach to feature discretization and selection

Many learning problems require handling high dimensional datasets with a relatively small number of instances. Learning algorithms are thus confronted with the curse of dimensionality, and need to address it in order to be effective. Examples of these types of data include the bag-of-words representation in text classification problems and gene expression data for tumor detection/classification. Usually, among the high number of features characterizing the instances, many may be irrelevant (or even detrimental) for the learning tasks. It is thus clear that there is a need for adequate techniques for feature representation, reduction, and selection, to improve both the classification accuracy and the memory requirements. In this paper, we propose combined unsupervised feature discretization and feature selection techniques, suitable for medium and high-dimensional datasets. The experimental results on several standard datasets, with both sparse and dense features, show the efficiency of the proposed techniques as well as improvements over previous related techniques.

Efficient feature selection filters for high-dimensional data

Feature selection is a central problem in machine learning and pattern recognition. On large datasets (in terms of dimension and/or number of instances), using search-based or wrapper techniques can be cornputationally prohibitive. Moreover, many filter methods based on relevance/redundancy assessment also take a prohibitively long time on high-dimensional. datasets. In this paper, we propose efficient unsupervised and supervised feature selection/ranking filters for high-dimensional datasets. These methods use low-complexity relevance and redundancy criteria, applicable to supervised, semi-supervised, and unsupervised learning, being able to act as pre-processors for computationally intensive methods to focus their attention on smaller subsets of promising features. The experimental results, with up to 10(5) features, show the time efficiency of our methods, with lower generalization error than state-of-the-art techniques, while being dramatically simpler and faster.

Plataforma de negociação de componentes multimédia

O objectivo da tese é demonstrar a adequação do paradigma dos mercados electrónicos baseados em agentes para transaccionar objectos multimédia em função do perfil dos espectadores. Esta dissertação descreve o projecto realizado no âmbito da plataforma de personalização de conteúdos em construção. O domínio de aplicação adoptado foi a personalização dos intervalos publicitários difundidos pelos distribuidores de conteúdos multimédia, i.e., pretende-se gerar em tempo útil o alinhamento de anúncios publicitários que melhor se adeqúe ao perfil de um espectador ou de um grupo de espectadores. O projecto focou-se no estudo e selecção das tecnologias de suporte, na concepção da arquitectura e no desenvolvimento de um protótipo que permitisse realizar diversas experiências nomeadamente com diferentes estratégias e tipos de mercado. A arquitectura proposta para a plataforma consiste num sistema multiagente organizado em três camadas que disponibiliza interfaces do tipo serviço Web com o exterior. A camada de topo é constituída por agentes de interface com o exterior. Na camada intermédia encontram-se os agentes autónomos que modelam as entidades produtoras e consumidoras de componentes multimédia assim como a entidade reguladora do mercado. Estes agentes registam-se num serviço de registo próprio onde especificam os componentes multimédia que pretendem negociar. Na camada inferior realiza-se o mercado que é constituído por agentes delegados dos agentes da camada superior. O lançamento do mercado é efectuado através de uma interface e consiste na escolha do tipo de mercado e no tipo de itens a negociar. Este projecto centrou-se na realização da camada do mercado e da parte da camada intermédia de apoio às actividades de negociação no mercado. A negociação é efectuada em relação ao preço da transmissão do anúncio no intervalo em preenchimento. Foram implementados diferentes perfis de negociação com tácticas, incrementos e limites de variação de preço distintos. Em termos de protocolos de negociação, adoptou-se uma variante do Iterated Contract Net – o Fixed Iterated Contract Net. O protótipo resultante foi testado e depurado com sucesso.

Hexose and pentose transport inascomycetousyeasts: an overview

FEMS Yeast Research, Vol. 9, nº 4

Otimização energética em redes óticas de próxima geração Flex-grid vs Fixed-grid

Trabalho Final de Mestrado para obtenção do grau de Mestre em Engenharia de Eletrónica e Telecomunicações

Geography of the Information Society

All every day activities take place in space. And it is upon this that all information and knowledge revolve. The latter are the key elements in the organisation of territories. Their creation, use and distribution should therefore occur in a balanced way throughout the whole territory in order to allow all individuals to participate in an egalitarian society, in which the flow of knowledge can take precedence over the flow of interests. The information society depends, to a large extent, on the technological capacity to disseminate information and, consequently, the knowledge throughout territory, thereby creating conditions which allow a more balanced development, from the both the social and economic points of view thus avoiding the existence of info-exclusion territories. Internet should therefore be considered more than a mere technology, given that its importance goes well beyond the frontiers of culture and society. It is already a part of daily life and of the new forms of thinking and transmitting information, thus making it a basic necessity essential, for a full socio-economic development. Its role as a platform of creation and distribution of content is regarded as an indispensable element for education in today’s society, since it makes information a much more easily acquired benefit.”…in the same way that the new technologies of generation and distribution of energy allowed factories and large companies to establish themselves as the organisational bases of industrial society, so the internet today constitutes the technological base of the organisational form that characterises the Information Era: the network” (CASTELLS, 2004:15). The changes taking place today in regional and urban structures are increasingly more evident due to a combination of factors such as faster means of transport, more efficient telecommunications and other cheaper and more advanced technologies of information and knowledge. Although their impact on society is obvious, society itself also has a strong influence on the evolution of these technologies. And although physical distance has lost much of the responsibility it had towards explaining particular phenomena of the economy and of society, other aspects such as telecommunications, new forms of mobility, the networks of innovation, the internet, cyberspace, etc., have become more important, and are the subject of study and profound analysis. The science of geographical information, allows, in a much more rigorous way, the analysis of problems thus integrating in a much more balanced way, the concepts of place, of space and of time. Among the traditional disciplines that have already found their place in this process of research and analysis, we can give special attention to a geography of new spaces, which, while not being a geography of ‘innovation’, nor of the ‘Internet’, nor even ‘virtual’, which can be defined as one of the ‘Information Society’, encompassing not only the technological aspects but also including a socio-economic approach. According to the last European statistical data, Portugal shows a deficit in terms of information and knowledge dissemination among its European partners. Some of the causes are very well identified - low levels of scholarship, weak investments on innovation and R&D (both private and public sector) - but others seem to be hidden behind socio-economical and technological factors. So, the justification of Portugal as the case study appeared naturally, on a difficult quest to find the major causes to territorial asymmetries. The substantial amount of data needed for this work was very difficult to obtain and for the islands of Madeira and Azores was insufficient, so only Continental Portugal was considered for this study. In an effort to understand the various aspects of the Geography of the Information Society and bearing in mind the increasing generalised use of information technologies together with the range of technologies available for the dissemination of information, it is important to: (i) Reflect on the geography of the new socio-technological spaces. (ii) Evaluate the potential for the dissemination of information and knowledge through the selection of variables that allow us to determine the dynamic of a given territory or region; (iii) Define a Geography of the Information Society in Continental Portugal.

Personalised placement in networked video

Personalised video can be achieved by inserting objects into a video play-out according to the viewer's profile. Content which has been authored and produced for general broadcast can take on additional commercial service features when personalised either for individual viewers or for groups of viewers participating in entertainment, training, gaming or informational activities. Although several scenarios and use-cases can be envisaged, we are focussed on the application of personalised product placement. Targeted advertising and product placement are currently garnering intense interest in the commercial networked media industries. Personalisation of product placement is a relevant and timely service for next generation online marketing and advertising and for many other revenue generating interactive services. This paper discusses the acquisition and insertion of media objects into a TV video play-out stream where the objects are determined by the profile of the viewer. The technology is based on MPEG-4 standards using object based video and MPEG-7 for metadata. No proprietary technology or protocol is proposed. To trade the objects into the video play-out, a Software-as-a-Service brokerage platform based on intelligent agent technology is adopted. Agencies, libraries and service providers are represented in a commercial negotiation to facilitate the contractual selection and usage of objects to be inserted into the video play-out.

CardioBengo study protocol: a population based cardiovascular longitudinal study in Bengo Province, Angola

Background: Cardiovascular diseases and other non-communicable diseases are major causes of morbidity and mortality, responsible for 38 million deaths in 2012, 75 % occurring in low- and middle-income countries. Most of these countries are facing a period of epidemiological transition, being confronted with an increased burden of non-communicable diseases, which challenge health systems mainly designed to deal with infectious diseases. With the adoption of the World Health Organization “Global Action Plan for the Prevention and Control of non-communicable diseases, 2013–2020”, the national dimension of risk factors for non-communicable diseases must be reported on a regular basis. Angola has no national surveillance system for non-communicable diseases, and periodic population-based studies can help to overcome this lack of information. CardioBengo will collect information on risk factors, awareness rates and prevalence of symptoms relevant to cardiovascular diseases, to assist decision makers in the implementation of prevention and treatment policies and programs. Methods: CardioBengo is designed as a research structure that comprises a cross-sectional component, providing baseline information and the assembling of a cohort to follow-up the dynamics of cardiovascular diseases risk factors in the catchment area of the Dande Health and Demographic Surveillance System of the Health Research Centre of Angola, in Bengo Province, Angola. The World Health Organization STEPwise approach to surveillance questionnaires and procedures will be used to collect information on a representative sex-age stratified sample, aged between 15 and 64 years old. Discussion: CardioBengo will recruit the first population cohort in Angola designed to evaluate cardiovascular diseases risk factors. Using the structures in place of the Dande Health and Demographic Surveillance System and a reliable methodology that generates comparable results with other regions and countries, this study will constitute a useful tool for the surveillance of cardiovascular diseases. Like all longitudinal studies, a strong concern exists regarding dropouts, but strategies like regular visits to selected participants and a strong community involvement are in place to minimize these occurrences.

The role of the efflux mechanisms in multidrug resistance in Mycobacterium tuberculosis

Abstract The emergence of multi and extensively drug resistant tuberculosis (MDRTB and XDRTB) has increased the concern of public health authorities around the world. The World Health Organization has defined MDRTB as tuberculosis (TB) caused by organisms resistant to at least isoniazid and rifampicin, the main first-line drugs used in TB therapy, whereas XDRTB refers to TB resistant not only to isoniazid and rifampicin, but also to a fluoroquinolone and to at least one of the three injectable second-line drugs, kanamycin, amikacin and capreomycin. Resistance in Mycobacterium tuberculosis is mainly due to the occurrence of spontaneous mutations and followed by selection of mutants by subsequent treatment. However, some resistant clinical isolates do not present mutations in any genes associated with resistance to a given antibiotic, which suggests that other mechanism(s) are involved in the development of drug resistance, namely the presence of efflux pump systems that extrude the drug to the exterior of the cell, preventing access to its target. Increased efflux activity can occur in response to prolonged exposure to subinhibitory concentrations of anti-TB drugs, a situation that may result from inadequate TB therapy. The inhibition of efflux activity with a non-antibiotic inhibitor may restore activity of an antibiotic subject to efflux and thus provide a way to enhance the activity of current anti-TB drugs. The work described in this thesis foccus on the study of efflux mechanisms in the development of multidrug resistance in M. tuberculosis and how phenotypic resistance, mediated by efflux pumps, correlates with genetic resistance. In order to accomplish this goal, several experimental protocols were developed using biological models such as Escherichia coli, the fast growing mycobacteria Mycobacterium smegmatis, and Mycobacterium avium, before their application to M. tuberculosis. This approach allowed the study of the mechanisms that result in the physiological adaptation of E. coli to subinhibitory concentrations of tetracycline (Chapter II), the development of a fluorometric method that allows the detection and quantification of efflux of ethidium bromide (Chapter III), the characterization of the ethidium bromide transport in M. smegmatis (Chapter IV) and the contribution of efflux activity to macrolide resistance in Mycobacterium avium complex (Chapter V). Finally, the methods developed allowed the study of the role of efflux pumps in M. tuberculosis strains induced to isoniazid resistance (Chapter VI). By this manner, in Chapter II it was possible to observe that the physiological adaptation of E. coli to tetracycline results from an interplay between events at the genetic level and protein folding that decrease permeability of the cell envelope and increase efflux pump activity. Furthermore, Chapter III describes the development of a semi-automated fluorometric method that allowed the correlation of this efflux activity with the transport kinetics of ethidium bromide (a known efflux pump substrate) in E. coli and the identification of efflux inhibitors. Concerning M. smegmatis, we have compared the wild-type M. smegmatis mc2155 with knockout mutants for LfrA and MspA for their ability to transport ethidium bromide. The results presented in Chapter IV showed that MspA, the major porin in M. smegmatis, plays an important role in the entrance of ethidium bromide and antibiotics into the cell and that efflux via the LfrA pump is involved in low-level resistance to these compounds in M. smegmatis. Chapter V describes the study of the contribution of efflux pumps to macrolide resistance in clinical M. avium complex isolates. It was demonstrated that resistance to clarithromycin was significantly reduced in the presence of efflux inhibitors such as thioridazine, chlorpromazine and verapamil. These same inhibitors decreased efflux of ethidium bromide and increased the retention of [14C]-erythromycin in these isolates. Finaly, the methods developed with the experimental models mentioned above allowed the study of the role of efflux pumps on M. tuberculosis strains induced to isoniazid resistance. This is described in Chapter VI of this Thesis, where it is demonstrated that induced resistance to isoniazid does not involve mutations in any of the genes known to be associated with isoniazid resistance, but an efflux system that is sensitive to efflux inhibitors. These inhibitors decreased the efflux of ethidium bromide and also reduced the minimum inhibitory concentration of isoniazid in these strains. Moreover, expression analysis showed overexpression of genes that code for efflux pumps in the induced strains relatively to the non-induced parental strains. In conclusion, the work described in this thesis demonstrates that efflux pumps play an important role in the development of drug resistance, namely in mycobacteria. A strategy to overcome efflux-mediated resistance may consist on the use of compounds that inhibit efflux activity, restoring the activity of antimicrobials that are efflux pump substrates, a useful approach particularly in TB where the most effective treatment regimens are becoming uneffective due to the increase of MDRTB/XDRTB.

Enhancing the selection of a model-based clustering with external categorical variables

In cluster analysis, it can be useful to interpret the partition built from the data in the light of external categorical variables which are not directly involved to cluster the data. An approach is proposed in the model-based clustering context to select a number of clusters which both fits the data well and takes advantage of the potential illustrative ability of the external variables. This approach makes use of the integrated joint likelihood of the data and the partitions at hand, namely the model-based partition and the partitions associated to the external variables. It is noteworthy that each mixture model is fitted by the maximum likelihood methodology to the data, excluding the external variables which are used to select a relevant mixture model only. Numerical experiments illustrate the promising behaviour of the derived criterion.

Materials selection for a set of multiple parts considering manufacturing costs and weight reduction with structural isoperformance using direct multisearch optimization

Materials selection is a matter of great importance to engineering design and software tools are valuable to inform decisions in the early stages of product development. However, when a set of alternative materials is available for the different parts a product is made of, the question of what optimal material mix to choose for a group of parts is not trivial. The engineer/designer therefore goes about this in a part-by-part procedure. Optimizing each part per se can lead to a global sub-optimal solution from the product point of view. An optimization procedure to deal with products with multiple parts, each with discrete design variables, and able to determine the optimal solution assuming different objectives is therefore needed. To solve this multiobjective optimization problem, a new routine based on Direct MultiSearch (DMS) algorithm is created. Results from the Pareto front can help the designer to align his/hers materials selection for a complete set of materials with product attribute objectives, depending on the relative importance of each objective.

The thickness of the transport layer in stratified geomorphic flows

River Flow 2010

An adaptive IoT management infrastructure for eco transport networks

The complexity associated with fast growing of B2B and the lack of a (complete) suite of open standards makes difficulty to maintain the underlying collaborative processes. Aligned to this challenge, this paper aims to be a contribution to an open architecture of logistics and transport processes management system. A model of an open integrated system is being defined as an open computational responsibility from the embedded systems (on-board) as well as a reference implementation (prototype) of a host system to validate the proposed open interfaces. Embedded subsystem can, natively, be prepared to cooperate with other on-board units and with IT-systems in an infrastructure commonly referred to as a center information system or back-office. In interaction with a central system the proposal is to adopt an open framework for cooperation where the embedded unit or the unit placed somewhere (land/sea) interacts in response to a set of implemented capabilities.

Exploiting the bin-class histograms for feature selection on discrete data

In machine learning and pattern recognition tasks, the use of feature discretization techniques may have several advantages. The discretized features may hold enough information for the learning task at hand, while ignoring minor fluctuations that are irrelevant or harmful for that task. The discretized features have more compact representations that may yield both better accuracy and lower training time, as compared to the use of the original features. However, in many cases, mainly with medium and high-dimensional data, the large number of features usually implies that there is some redundancy among them. Thus, we may further apply feature selection (FS) techniques on the discrete data, keeping the most relevant features, while discarding the irrelevant and redundant ones. In this paper, we propose relevance and redundancy criteria for supervised feature selection techniques on discrete data. These criteria are applied to the bin-class histograms of the discrete features. The experimental results, on public benchmark data, show that the proposed criteria can achieve better accuracy than widely used relevance and redundancy criteria, such as mutual information and the Fisher ratio.