Modelo de data mining para detecção de tumores em exames de rastreio

Dissertação para obtenção do grau de Mestre em Engenharia Informática

Unmixing hyperspectral data: independent and dependent component analysis

The development of high spatial resolution airborne and spaceborne sensors has improved the capability of ground-based data collection in the fields of agriculture, geography, geology, mineral identification, detection [2, 3], and classification [4–8]. The signal read by the sensor from a given spatial element of resolution and at a given spectral band is a mixing of components originated by the constituent substances, termed endmembers, located at that element of resolution. This chapter addresses hyperspectral unmixing, which is the decomposition of the pixel spectra into a collection of constituent spectra, or spectral signatures, and their corresponding fractional abundances indicating the proportion of each endmember present in the pixel [9, 10]. Depending on the mixing scales at each pixel, the observed mixture is either linear or nonlinear [11, 12]. The linear mixing model holds when the mixing scale is macroscopic [13]. The nonlinear model holds when the mixing scale is microscopic (i.e., intimate mixtures) [14, 15]. The linear model assumes negligible interaction among distinct endmembers [16, 17]. The nonlinear model assumes that incident solar radiation is scattered by the scene through multiple bounces involving several endmembers [18]. Under the linear mixing model and assuming that the number of endmembers and their spectral signatures are known, hyperspectral unmixing is a linear problem, which can be addressed, for example, under the maximum likelihood setup [19], the constrained least-squares approach [20], the spectral signature matching [21], the spectral angle mapper [22], and the subspace projection methods [20, 23, 24]. Orthogonal subspace projection [23] reduces the data dimensionality, suppresses undesired spectral signatures, and detects the presence of a spectral signature of interest. The basic concept is to project each pixel onto a subspace that is orthogonal to the undesired signatures. As shown in Settle [19], the orthogonal subspace projection technique is equivalent to the maximum likelihood estimator. This projection technique was extended by three unconstrained least-squares approaches [24] (signature space orthogonal projection, oblique subspace projection, target signature space orthogonal projection). Other works using maximum a posteriori probability (MAP) framework [25] and projection pursuit [26, 27] have also been applied to hyperspectral data. In most cases the number of endmembers and their signatures are not known. Independent component analysis (ICA) is an unsupervised source separation process that has been applied with success to blind source separation, to feature extraction, and to unsupervised recognition [28, 29]. ICA consists in finding a linear decomposition of observed data yielding statistically independent components. Given that hyperspectral data are, in given circumstances, linear mixtures, ICA comes to mind as a possible tool to unmix this class of data. In fact, the application of ICA to hyperspectral data has been proposed in reference 30, where endmember signatures are treated as sources and the mixing matrix is composed by the abundance fractions, and in references 9, 25, and 31–38, where sources are the abundance fractions of each endmember. In the first approach, we face two problems: (1) The number of samples are limited to the number of channels and (2) the process of pixel selection, playing the role of mixed sources, is not straightforward. In the second approach, ICA is based on the assumption of mutually independent sources, which is not the case of hyperspectral data, since the sum of the abundance fractions is constant, implying dependence among abundances. This dependence compromises ICA applicability to hyperspectral images. In addition, hyperspectral data are immersed in noise, which degrades the ICA performance. IFA [39] was introduced as a method for recovering independent hidden sources from their observed noisy mixtures. IFA implements two steps. First, source densities and noise covariance are estimated from the observed data by maximum likelihood. Second, sources are reconstructed by an optimal nonlinear estimator. Although IFA is a well-suited technique to unmix independent sources under noisy observations, the dependence among abundance fractions in hyperspectral imagery compromises, as in the ICA case, the IFA performance. Considering the linear mixing model, hyperspectral observations are in a simplex whose vertices correspond to the endmembers. Several approaches [40–43] have exploited this geometric feature of hyperspectral mixtures [42]. Minimum volume transform (MVT) algorithm [43] determines the simplex of minimum volume containing the data. The MVT-type approaches are complex from the computational point of view. Usually, these algorithms first find the convex hull defined by the observed data and then fit a minimum volume simplex to it. Aiming at a lower computational complexity, some algorithms such as the vertex component analysis (VCA) [44], the pixel purity index (PPI) [42], and the N-FINDR [45] still find the minimum volume simplex containing the data cloud, but they assume the presence in the data of at least one pure pixel of each endmember. This is a strong requisite that may not hold in some data sets. In any case, these algorithms find the set of most pure pixels in the data. Hyperspectral sensors collects spatial images over many narrow contiguous bands, yielding large amounts of data. For this reason, very often, the processing of hyperspectral data, included unmixing, is preceded by a dimensionality reduction step to reduce computational complexity and to improve the signal-to-noise ratio (SNR). Principal component analysis (PCA) [46], maximum noise fraction (MNF) [47], and singular value decomposition (SVD) [48] are three well-known projection techniques widely used in remote sensing in general and in unmixing in particular. The newly introduced method [49] exploits the structure of hyperspectral mixtures, namely the fact that spectral vectors are nonnegative. The computational complexity associated with these techniques is an obstacle to real-time implementations. To overcome this problem, band selection [50] and non-statistical [51] algorithms have been introduced. This chapter addresses hyperspectral data source dependence and its impact on ICA and IFA performances. The study consider simulated and real data and is based on mutual information minimization. Hyperspectral observations are described by a generative model. This model takes into account the degradation mechanisms normally found in hyperspectral applications—namely, signature variability [52–54], abundance constraints, topography modulation, and system noise. The computation of mutual information is based on fitting mixtures of Gaussians (MOG) to data. The MOG parameters (number of components, means, covariances, and weights) are inferred using the minimum description length (MDL) based algorithm [55]. We study the behavior of the mutual information as a function of the unmixing matrix. The conclusion is that the unmixing matrix minimizing the mutual information might be very far from the true one. Nevertheless, some abundance fractions might be well separated, mainly in the presence of strong signature variability, a large number of endmembers, and high SNR. We end this chapter by sketching a new methodology to blindly unmix hyperspectral data, where abundance fractions are modeled as a mixture of Dirichlet sources. This model enforces positivity and constant sum sources (full additivity) constraints. The mixing matrix is inferred by an expectation-maximization (EM)-type algorithm. This approach is in the vein of references 39 and 56, replacing independent sources represented by MOG with mixture of Dirichlet sources. Compared with the geometric-based approaches, the advantage of this model is that there is no need to have pure pixels in the observations. The chapter is organized as follows. Section 6.2 presents a spectral radiance model and formulates the spectral unmixing as a linear problem accounting for abundance constraints, signature variability, topography modulation, and system noise. Section 6.3 presents a brief resume of ICA and IFA algorithms. Section 6.4 illustrates the performance of IFA and of some well-known ICA algorithms with experimental data. Section 6.5 studies the ICA and IFA limitations in unmixing hyperspectral data. Section 6.6 presents results of ICA based on real data. Section 6.7 describes the new blind unmixing scheme and some illustrative examples. Section 6.8 concludes with some remarks.

Agent Inferencing Meets the Semantic Web

We provide all agent; the capability to infer the relations (assertions) entailed by the rules that, describe the formal semantics of art RDFS knowledge-base. The proposed inferencing process formulates each semantic restriction as a rule implemented within a, SPARQL query statement. The process expands the original RDF graph into a fuller graph that. explicitly captures the rule's described semantics. The approach is currently being explored in order to support descriptions that follow the generic Semantic Web Rule Language. An experiment, using the Fire-Brigade domain, a small-scale knowledge-base, is adopted to illustrate the agent modeling method and the inferencing process.

The problem of estimating the volatility of zero coupon bond interest rate

Financial literature and financial industry use often zero coupon yield curves as input for testing hypotheses, pricing assets or managing risk. They assume this provided data as accurate. We analyse implications of the methodology and of the sample selection criteria used to estimate the zero coupon bond yield term structure on the resulting volatility of spot rates with different maturities. We obtain the volatility term structure using historical volatilities and Egarch volatilities. As input for these volatilities we consider our own spot rates estimation from GovPX bond data and three popular interest rates data sets: from the Federal Reserve Board, from the US Department of the Treasury (H15), and from Bloomberg. We find strong evidence that the resulting zero coupon bond yield volatility estimates as well as the correlation coefficients among spot and forward rates depend significantly on the data set. We observe relevant differences in economic terms when volatilities are used to price derivatives.

Fungal contamination of poultries litter: a public health problem

Exposure to certain fungi can cause human illness. Fungi cause adverse human health effects through three specific mechanisms: generation of a harmful immune response (e.g., allergy or hypersensitivity pneumonitis); direct infection by the fungal organism; by toxic-irritant effects from mold byproducts, such as mycotoxins. In Portugal there is an increasingly industry of large facilities that produce whole chickens for domestic consumption and only few investigations have reported on fungal contamination of the poultry litter. The material used for poultry litter is varied but normally can be constitute by: pine shavings; sawdust of eucalyptus; other types of wood; peanut; coffee; sugar cane; straw; hay; grass; paper processed. Litter is one of the most contributive factors to fungal contamination in poultries. Spreading litter is one of the tasks that normally involve higher exposure of the poultry workers to dust, fungi and their metabolites, such as VOC’s and mycotoxins. After being used and removed from poultries, litter is ploughed into agricultural soils, being this practice potentially dangerous for the soil environment, as well for both humans and animals. The goal of this study was to characterize litter’s fungal contamination and also to report the incidence of keratinophilic and toxigenic fungi.

Positive Solutions of the Dirichlet Problem for the One-dimensional Minkowski-Curvature Equation

We discuss existence and multiplicity of positive solutions of the Dirichlet problem for the quasilinear ordinary differential equation-(u' / root 1 - u'(2))' = f(t, u). Depending on the behaviour of f = f(t, s) near s = 0, we prove the existence of either one, or two, or three, or infinitely many positive solutions. In general, the positivity of f is not required. All results are obtained by reduction to an equivalent non-singular problem to which variational or topological methods apply in a classical fashion.

Fungal contamination of poultry litter: a public health problem

Although numerous studies have been conducted on microbial contaminants associated with various stages related to poultry and meat products processing, only a few reported on fungal contamination of poultry litter. The goals of this study were to (1) characterize litter fungal contamination and (2) report the incidence of keratinophilic and toxigenic fungi presence. Seven fresh and 14 aged litter samples were collected from 7 poultry farms. In addition, 27 air samples of 25 litters were also collected through impaction method, and after laboratory processing and incubation of collected samples, quantitative colony-forming units (CFU/m3) and qualitative results were obtained. Twelve different fungal species were detected in fresh litter and Penicillium was the most frequent genus found (59.9%), followed by Alternaria (17.8%), Cladosporium (7.1%), and Aspergillus (5.7%). With respect to aged litter, 19 different fungal species were detected, with Penicillium sp. the most frequently isolated (42.3%), followed by Scopulariopsis sp. (38.3%), Trichosporon sp. (8.8%), and Aspergillus sp. (5.5%). A significant positive correlation was found between litter fungal contamination (CFU/g) and air fungal contamination (CFU/m3). Litter fungal quantification and species identification have important implications in the evaluation of potential adverse health risks to exposed workers and animals. Spreading of poultry litter in agricultural fields is a potential public health concern, since keratinophilic (Scopulariopsis and Fusarium genus) as well as toxigenic fungi (Aspergillus, Fusarium, and Penicillium genus) were isolated.

Biblioteca de navegação autónoma baseada em campos de potencial

Trabalho de Final de Mestrado para obtenção do grau de Mestre em Engenharia Informática e de Computadores

Between theory and practice – teaching maths to elementary school students in Poland

The development of children's school achievements in mathematics is one of the most important aims of education in Poland. The results of research concerning monitoring of school achievements in maths is not optimistic. We can observe low levels of children’s understanding of the merits of maths, self-developed strategies in solving problems and practical usage of maths skills. This article frames the discussion of this problem in its psychological and didactic context and analyses the causes as they relate to school practice in teaching maths

A remote virtual experiment in mechanical engineering

Conferência: 2nd Experiment at International Conference (Exp at)- Univ Coimbra, Coimbra, Portugal - Sep 18-20, 2013

Fast growing fungi: a problem to be solved to achieve characterization of occupational exposure to fungi in cork industry

Chrysonilia sitophila is a common mould in cork industry and has been identified as a cause of IgE sensitization and occupational asthma. This fungal species have a fast growth rate that may inhibit others species’ growth causing underestimated data from characterization of occupational fungal exposure. Aiming to ascertain occupational exposure to fungi in cork industry, were analyzed papers from 2000 about the best air sampling method, to obtain quantification and identification of all airborne culturable fungi, besides the ones that have fast-growing rates. Impaction method don’t allows the collection of a representative air volume, because even with some media that restricts the growth of the colonies, in environments with higher fungal load, such as cork industry, the counting of the colonies is very difficult. Otherwise, impinger method permits the collection of a representative air volume, since we can make dilution of the collected volume. Besides culture methods that allows fungal identification trough macro- and micro-morphology, growth features, thermotolerance and ecological data, we can apply molecular biology with the impinger method, to detect the presence of non-viable particles and potential mycotoxin producers’ strains, and also to detect mycotoxins presence with ELISA or HPLC. Selection of the best air sampling method in each setting is crucial to achieve characterization of occupational exposure to fungi. Information about the prevalent fungal species in each setting and also the eventual fungal load it’s needed for a criterious selection.

Feature selection for clustering categorical data with an embedded modelling approach

Research on the problem of feature selection for clustering continues to develop. This is a challenging task, mainly due to the absence of class labels to guide the search for relevant features. Categorical feature selection for clustering has rarely been addressed in the literature, with most of the proposed approaches having focused on numerical data. In this work, we propose an approach to simultaneously cluster categorical data and select a subset of relevant features. Our approach is based on a modification of a finite mixture model (of multinomial distributions), where a set of latent variables indicate the relevance of each feature. To estimate the model parameters, we implement a variant of the expectation-maximization algorithm that simultaneously selects the subset of relevant features, using a minimum message length criterion. The proposed approach compares favourably with two baseline methods: a filter based on an entropy measure and a wrapper based on mutual information. The results obtained on synthetic data illustrate the ability of the proposed expectation-maximization method to recover ground truth. An application to real data, referred to official statistics, shows its usefulness.

Corrosão das armaduras do betão armado: causas, consequências, prevenção e projeto de durabilidade

Trabalho Final de Mestrado para obtenção do grau de Mestre em Engenharia Civil

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.

Theory and phenomenology of two-higgs-doublet models

We discuss theoretical and phenomenological aspects of two-Higgs-doublet extensions of the Standard Model. In general, these extensions have scalar mediated flavour changing neutral currents which are strongly constrained by experiment. Various strategies are discussed to control these flavour changing scalar currents and their phenomenological consequences are analysed. In particular, scenarios with natural flavour conservation are investigated, including the so-called type I and type II models as well as lepton-specific and inert models. Type III models are then discussed, where scalar flavour changing neutral currents are present at tree level, but are suppressed by either a specific ansatz for the Yukawa couplings or by the introduction of family symmetries leading to a natural suppression mechanism. We also consider the phenomenology of charged scalars in these models. Next we turn to the role of symmetries in the scalar sector. We discuss the six symmetry-constrained scalar potentials and their extension into the fermion sector. The vacuum structure of the scalar potential is analysed, including a study of the vacuum stability conditions on the potential and the renormalization-group improvement of these conditions is also presented. The stability of the tree level minimum of the scalar potential in connection with electric charge conservation and its behaviour under CP is analysed. The question of CP violation is addressed in detail, including the cases of explicit CP violation and spontaneous CP violation. We present a detailed study of weak basis invariants which are odd under CP. These invariants allow for the possibility of studying the CP properties of any two-Higgs-doublet model in an arbitrary Higgs basis. A careful study of spontaneous CP violation is presented, including an analysis of the conditions which have to be satisfied in order for a vacuum to violate CP. We present minimal models of CP violation where the vacuum phase is sufficient to generate a complex CKM matrix, which is at present a requirement for any realistic model of spontaneous CP violation.