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. © 2014 Springer-Verlag Berlin Heidelberg.

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.

Synthesis of titanate nanostructures using amorphous precursor material and their adsorption/photocatalytic properties

This article reports on a new and swift hydrothermal chemical route to prepare titanate nanostructures (TNS) avoiding the use of crystalline TiO2 as starting material. The synthesis approach uses a commercial solution of TiCl3 as titanium source to prepare an amorphous precursor, circumventing the use of hazardous chemical compounds. The influence of the reaction temperature and dwell autoclave time on the structure and morphology of the synthesised materials was studied. Homogeneous titanate nanotubes with a high length/diameter aspect ratio were synthesised at 160 degrees C and 24 h. A band gap of 3.06 +/- 0.03 eV was determined for the TNS samples prepared in these experimental conditions. This value is red shifted by 0.14 eV compared to the band gap value usually reported for the TiO2 anatase. Moreover, such samples show better adsorption capacity and photocatalytic performance on the dye rhodamine 6G (R6G) photodegradation process than TiO2 nanoparticles. A 98% reduction of the R6G concentration was achieved after 45 min of irradiation of a 10 ppm dye aqueous solution and 1 g L-1 of TNS catalyst.

Caracterização mecânica e análise da resistência à corrosão em diferentes solos, de aços inoxidáveis duplex

O aço inoxidável é um material de alta durabilidade e resistência, utilizado nos mais diversos setores comerciais. O conhecimento das suas características e propriedades é essencial para uma escolha mais seletiva e vantajosa. Pretende‐se com este trabalho, estudar as propriedades mecânicas, a resistência ao desgaste e o comportamento, em ambientes mais agressivos de três tipos de aços inoxidáveis duplex, produzidos pela OUTOKUMPU e comercialmente conhecidos pelas designações LDX 2101, SAF 2507 e o SAF 2205. Para tal efeito foram realizados ensaios de Dureza Vickers, antes e após dobragem; ensaios de tracção em duas diferentes situações: seguindo a norma NP EN 10002‐ 1:2006 e após a realização de alguns ciclos de carga e descarga; ensaio de Impacto à temperatura ambiente e após arrefecimento criogénico; ensaio de resistência à corrosão. Também foi feito o estudo de resistência ao desgaste com base na técnica de micro‐abrasão por esfera rotativa e o estudo metalográfico. Foi também objetivo deste trabalho, relacionar o uso dos referidos aços duplex e as consequências que esse uso tem para o meio ambiente, bem como o seu comportamento quando exposto a condições extremas. Para tal, foram colocadas amostras dos referidos aços, em dois tipos de solos com condições de humidade e acidez diferentes analisando o seu estado após imersão em solo com condições controladas durante 272 dias.

Desenvolvimento de um modelo de procurement e abastecimento para uma união de mutualistas

Mestrado em Engenharia Mecânica – Gestão Industrial

Problemática associada à utilização de diferentes tipos de mochila para transporte de material escolar

O transporte de cargas é uma tarefa comum para crianças, adolescentes e adultos, pela necessidade de transferência diária de objetos pessoais, livros e artigos de papelaria para os locais de trabalho ou escolas. Diversos autores apontam que o peso carregado durante transporte de material é o principal responsável pelo aparecimento de dor lombar. Deste modo é importante o constante estudo da temática para a definição recomendações e limites. O presente estudo teve como principais objetivos a caraterização da problemática associada à utilização de mochilas e a determinação do Peso Máximo Aceitável (PMA) e do Índice de Esforço Percebido (IEP) para a tarefa de transporte de mochilas, através da abordagem psicofísica. O estudo foi desenvolvido com estudantes do 7º, 8º e 9º ano de escolaridade e, foi dividido em duas fases. Na 1ª fase foram aplicados questionários para a análise da problemática associada à utilização de diferentes tipos de mochilas escolares. Nesta fase, foram incluídos aspetos associados à identificação do tipo de mochila mais utilizada, as rotinas e hábitos dos estudantes e as características da mochila utilizada. Verificou-se que os estudantes utilizam, maioritariamente, a mochila de duas alças para transporte de material escolar. Posteriormente foram efetuadas medições de peso da mochila, altura e peso aos 131 estudantes que constituíram a amostra da 1º fase. O principal objetivo deste ponto foi identificar o tipo de mochila habitualmente utilizada pelos estudantes assim como, o peso transportado nas mochilas. Na 2ª fase foi efetuado um estudo para a determinação do PMA e do IEP, através da abordagem psicofísica, para a tarefa de transporte de mochila, considerando-se uma amostra constituída por 10 estudantes. Para este estudo, apenas foi considerada a mochila mais frequentemente utilizada, identificada na 1º fase. A tarefa consistiu no transporte da mochila nos dois ombros e com as alças devidamente ajustadas ao corpo, num percurso pré-definido, de acordo com o procedimento experimental. Os resultados indicaram que nem todos os estudantes transportam mochilas com pesos dentro das recomendações da Organização Mundial de Saúde. O PMA determinado pelos estudantes foi de 6.8 kg para a mochila de duas alças e a região dos ombros foi identificada durante todo o estudo como sendo a que apresentava maior intensidade de dor durante o transporte da mochila.

GFRP waste reinforced polymer mortars: a new waste management solution for GFRP scrap material

To date, glass fibre reinforced polymer (GFRP) waste recycling is very limited and restricted by thermoset nature of binder matrix and lack of economically viable enduse applications for the recyclates. In this study, efforts were made in order to recycle grinded GFRP waste proceeding from pultrusion production scrap, into new and sustainable composite materials. For this purpose, GFRP waste recyclates, a mix of powdered and fibrous materials, were incorporated into polyester based mortars as fine aggregate and filler replacements, at different load contents (between 4% up to 12% of total mass) and particle size distributions. Potential recycling solution was assessed by mechanical behaviour of resultant GFRP waste modified polymer mortars. Test results revealed that GFRP waste filled polymer mortars present improved flexural and compressive behaviour over unmodified polyester based mortars, thus indicating the feasibility of GFRP waste reuse in concrete-polymer composites.

An experimental study on recycling and reuse of pultruded composite material wastes

Glass fibre-reinforced plastics (GFRP) have been considered inherently difficult to recycle due to both: cross-linked nature of thermoset resins, which cannot be remoulded, and complex composition of the composite itself. Presently, most of the GFRP waste is landfilled leading to negative environmental impacts and supplementary added costs. With an increasing awareness of environmental matters and the subsequent desire to save resources, recycling would convert an expensive waste disposal into a profitable reusable material. In this study, efforts were made in order to recycle grinded GFRP waste, proceeding from pultrusion production scrap, into new and sustainable composite materials. For this purpose, GFRP waste recyclates, were incorporated into polyester based mortars as fine aggregate and filler replacements at different load contents and particle size distributions. Potential recycling solution was assessed by mechanical behaviour of resultant GFRP waste modified polymer mortars. Results revealed that GFRP waste filled polymer mortars present improved flexural and compressive behavior over unmodified polyester based mortars, thus indicating the feasibility of the GFRP industrial waste reuse into concrete-polymer composite materials.

Dynamic behaviour of soft core sandwich beam structures using kriging-based layerwise models

Sandwich structures with soft cores are widely used in applications where a high bending stiffness is required without compromising the global weight of the structure, as well as in situations where good thermal and damping properties are important parameters to observe. As equivalent single layer approaches are not the more adequate to describe realistically the kinematics and the stresses distributions as well as the dynamic behaviour of this type of sandwiches, where shear deformations and the extensibility of the core can be very significant, layerwise models may provide better solutions. Additionally and in connection with this multilayer approach, the selection of different shear deformation theories according to the nature of the material that constitutes the core and the outer skins can predict more accurately the sandwich behaviour. In the present work the authors consider the use of different shear deformation theories to formulate different layerwise models, implemented through kriging-based finite elements. The viscoelastic material behaviour, associated to the sandwich core, is modelled using the complex approach and the dynamic problem is solved in the frequency domain. The outer elastic layers considered in this work may also be made from different nanocomposites. The performance of the models developed is illustrated through a set of test cases. (C) 2015 Elsevier Ltd. All rights reserved.

O outro lado da solidão: trabalho de projecto

Projecto elaborado com vista à obtenção do Grau de Mestre em Teatro, área de especialização em Artes Performativas – Interpretação.

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.

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.

Dual-phase inkjet printed electrochromic layers based on PTA and WOX/TiO2 nanoparticles for electrochromic applications

Thesis submitted in Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa for the degree of Master in Materials Engineering

The selection of senior civil servants in the context of the different public administration models: an issue of accountability

Comunicação apresentada no Congresso do IIAS-IISA no âmbito do IX Grupo de Estudo: Serviço público e política, realizado em Ifrane, Marrocos de 13 a 17 de junho de 2014