Utilizando Pesos estáticos e dinâmicos em sistemas multi-classificadores com diferentes níveis de diversidade

Although some individual techniques of supervised Machine Learning (ML), also known as classifiers, or algorithms of classification, to supply solutions that, most of the time, are considered efficient, have experimental results gotten with the use of large sets of pattern and/or that they have a expressive amount of irrelevant data or incomplete characteristic, that show a decrease in the efficiency of the precision of these techniques. In other words, such techniques can t do an recognition of patterns of an efficient form in complex problems. With the intention to get better performance and efficiency of these ML techniques, were thought about the idea to using some types of LM algorithms work jointly, thus origin to the term Multi-Classifier System (MCS). The MCS s presents, as component, different of LM algorithms, called of base classifiers, and realized a combination of results gotten for these algorithms to reach the final result. So that the MCS has a better performance that the base classifiers, the results gotten for each base classifier must present an certain diversity, in other words, a difference between the results gotten for each classifier that compose the system. It can be said that it does not make signification to have MCS s whose base classifiers have identical answers to the sames patterns. Although the MCS s present better results that the individually systems, has always the search to improve the results gotten for this type of system. Aim at this improvement and a better consistency in the results, as well as a larger diversity of the classifiers of a MCS, comes being recently searched methodologies that present as characteristic the use of weights, or confidence values. These weights can describe the importance that certain classifier supplied when associating with each pattern to a determined class. These weights still are used, in associate with the exits of the classifiers, during the process of recognition (use) of the MCS s. Exist different ways of calculating these weights and can be divided in two categories: the static weights and the dynamic weights. The first category of weights is characterizes for not having the modification of its values during the classification process, different it occurs with the second category, where the values suffers modifications during the classification process. In this work an analysis will be made to verify if the use of the weights, statics as much as dynamics, they can increase the perfomance of the MCS s in comparison with the individually systems. Moreover, will be made an analysis in the diversity gotten for the MCS s, for this mode verify if it has some relation between the use of the weights in the MCS s with different levels of diversity

Caracterização dos enunciados de um aluno não-falante usuário de recurso suplementar de comunicação durante a construção de histórias

A narrativa pode conectar elementos diversos e, assim, promover a construção de sentenças verticais e horizontais, frente à mediação de interlocutor. Esta pesquisa visou analisar a extensão e a organização sintática dos enunciados de narrativas construídas por um aluno não-falante usuário de recurso suplementar de comunicação. Previamente à construção das narrativas, selecionadas de acordo com os centros de interesse, idade, grau de escolaridade e temáticas trabalhadas na escola do aluno, símbolos gráficos do Picture Communication Symbols eram confeccionados e adaptados para que, por meio destes e dos recursos verbais e não-verbais inerentes à atividade discursiva, a história pudesse ser contada, por acesso visual e auditivo, pelo interlocutor e recontada pelo aluno, após a estruturação e organização dos múltiplos signos em sentenças. Todas as interações foram filmadas e complementadas com anotações de um diário de registro contínuo. Após a análise das transcrições das fitas e das anotações do diário, foi possível determinar três estágios de unidades significativas: 1) Caracterização ou natureza dos elementos constituintes dos enunciados das narrativas em: a) objeto; b) símbolo gráfico; c) escrita; d) vocalização; e) gesto representativo; 2) Extensão dos enunciados das narrativas de 1 a 6 elementos, linearmente, e 7 ou mais elementos; 3) Organização sintática dos elementos constituintes das narrativas em sujeito, verbo e objeto isolados, em sentenças verticais, e associados, em sentenças horizontais. Ressaltou-se, assim, a ampliação e aprimoramento da organização e da estrutura sintática dos enunciados das narrativas construídas pelo aluno mediado pelo interlocutor.

Analysis of speech fluency in Williams syndrome

Williams syndrome (WS) is a neurodevelopmental genetic disorder, often referred as being characterized by dissociation between verbal and non-verbal abilities, although the number of studies disputing this proposal is emerging. Indeed, although they have been traditionally reported as displaying increased speech fluency, this topic has not been fully addressed in research. In previous studies carried out with a small group of individuals with WS, we reported speech breakdowns during conversational and autobiographical narratives suggestive of language difficulties. In the current study, we characterized the speech fluency profile using an ecologically based measure - a narrative task (story generation) was collected from a group of individuals with WS (n = 30) and typically developing group (n = 39) matched in mental age. Oral narratives were elicited using a picture stimulus - the cookie theft picture from Boston Diagnosis Aphasia Test. All narratives were analyzed according to typology and frequency of fluency breakdowns (non-stuttered and stuttered disfluencies). Oral narratives in WS group differed from typically developing group, mainly due to a significant increase in the frequency of disfluencies, particularly in terms of hesitations, repetitions and pauses. This is the first evidence of disfluencies in WS using an ecologically based task (oral narrative task), suggesting that these speech disfluencies may represent a significant marker of language problems in WS. (C) 2011 Elsevier Ltd. All rights reserved.

Desempenho de escolares com distúrbio de aprendizagem e dislexia em testes de processamento auditivo

OBJETIVO: caracterizar e comparar, por meio de testes comportamentais, o processamento auditivo de escolares com diagnóstico interdisciplinar de (I) distúrbio da aprendizagem, (II) dislexia e (III) escolares com bom desempenho acadêmico. MÉTODOS: participaram deste estudo 30 escolares na faixa etária de 8 a 16 anos de idade, de ambos os gêneros, de 2ª a 4ª séries do ensino fundamental, divididos em três grupos: GI composto por 10 escolares com diagnóstico interdisciplinar de distúrbio de aprendizagem, GII: composto por 10 escolares com diagnóstico interdisciplinar de dislexia e GIII composto por 10 escolares sem dificuldades de aprendizagem, pareados segundo gênero e faixa etária com os grupos GI e GII. Foram realizadas avaliação audiológica e de processamento auditivo. RESULTADOS: os escolares de GIII apresentaram desempenho superior nos testes de processamento auditivo em relação aos escolares de GI e GII. GI apresentou desempenho inferior nas habilidades auditivas avaliadas para testes dicóticos de dígitos e dissílabos alternados, logoaudiometria pediátrica, localização sonora, memória verbal e não-verbal, ao passo que GII apresentou as mesmas alterações de GI, com exceção do teste de logoaudiometria pediátrica. CONCLUSÃO: os escolares com transtornos de aprendizagem apresentaram desempenho inferior nos testes de processamento auditivo, sendo que os escolares com distúrbio de aprendizagem apresentaram maior número de habilidades auditivas alteradas, em comparação com os escolares com dislexia, por terem apresentado atenção sustentada reduzida. O grupo de escolares com dislexia apresentou alterações decorrentes da dificuldade relacionada à codificação e decodificação de estímulos sonoros.

Propriedades estáticas e dinâmicas de multicamadas magnéticas acopladas quasiperiodicamente

In this work we study, for two different growth directions, multilayers of nanometric magnetic metallic lms grown, using Fibonacci sequences, in such a way that the thickness of the non-magnetic spacer may vary from a pair of lms to another. We applied a phenomenological theory that uses the magnetic energy to describe the behavior of the system. After we found numerically the global minimum of the total energy, we used the equilibrium angles to obtain magnetization and magnetoresistance curves. Next, we solved the equation of motion of the multilayers to nd the dispersion relation for the system. The results show that, when spacers are used with thickness so that the biquadratic coupling is strong in comparison to the bilinear one, non usual behaviors for both magnetization and magnetoresistance are observed. For example, a dependence on the parity of the Fibonacci generation utilized for constructing the system, a low magnetoresistance step in low external magnetic fields and regions that show high sensibility to small variations of the applied field. Those behaviors are not present in quasiperiodic magnetic multilayers with constant spacer thickness

In silico network topology-based prediction of gene essentiality

The identification of genes essential for survival is important for the understanding of the minimal requirements for cellular life and for drug design. As experimental studies with the purpose of building a catalog of essential genes for a given organism are time-consuming and laborious, a computational approach which could predict gene essentiality with high accuracy would be of great value. We present here a novel computational approach, called NTPGE (Network Topology-based Prediction of Gene Essentiality), that relies on the network topology features of a gene to estimate its essentiality. The first step of NTPGE is to construct the integrated molecular network for a given organism comprising protein physical, metabolic and transcriptional regulation interactions. The second step consists in training a decision-tree-based machine-learning algorithm on known essential and non-essential genes of the organism of interest, considering as learning attributes the network topology information for each of these genes. Finally, the decision-tree classifier generated is applied to the set of genes of this organism to estimate essentiality for each gene. We applied the NTPGE approach for discovering the essential genes in Escherichia coli and then assessed its performance. (C) 2007 Elsevier B.V. All rights reserved.

Towards the prediction of essential genes by integration of network topology, cellular localization and biological process information

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Particle Competition and Cooperation in Networks for Semi-Supervised Learning

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Efeitos de modelagem do comportamento verbal e de instruções sobre o comportamento de encaixar peças

Pesquisas têm demonstrado diferenças nos efeitos dos comportamentos verbais modelado e instruído sobre o comportamento verbal e não-verbal. Este estudo investigou efeitos da modelagem do comportamento verbal e das instruções sobre o comportamento verbal (falar sobre encaixar peças) e o não-verbal (encaixar peças azuis e vermelhas, grandes e pequenas e quadradas e circulares) de 10 crianças, entre 8 e 9 anos de idade. A coleta de dados foi realizada em duas condições com cinco participantes. Condição 1: modelagem do comportamento verbal. Condição 2: apresentação de instruções para o comportamento não-verbal. Quando ocorreu a modelagem do comportamento verbal foram observadas mudanças correspondentes no comportamento não-verbal. As instruções produziram imediata adesão do comportamento não-verbal e, na seqüência, o desempenho foi alterado. Esses dados reafirmam a importância de ampliar o conhecimento dos efeitos da modelagem do comportamento verbal e das instruções sobre o comportamento de crianças em jogos como o utilizado nesta pesquisa.

Non isothermal study of the devitrification kinetics of fluoroindate glasses

The kinetics of crystallization in an indium fluoride-based glass was studied by a non-isothermal method using differential scanning calorimetry. The experiments led to an Avrami's exponent of 4.6 for solid glass and 2.2 for a powdered sample. The apparent activation energy for crystallization was found to be 130 kJ/mol for solid glass and 354 kJ/mol for the powder. These results express the profound effect of glass particle size on those kinetic parameters, as different crystallization mechanisms take place during sample heating.

Conceptual model for adaptable and extensible visual data exploration

Interactive visual representations complement traditional statistical and machine learning techniques for data analysis, allowing users to play a more active role in a knowledge discovery process and making the whole process more understandable. Though visual representations are applicable to several stages of the knowledge discovery process, a common use of visualization is in the initial stages to explore and organize a sometimes unknown and complex data set. In this context, the integrated and coordinated - that is, user actions should be capable of affecting multiple visualizations when desired - use of multiple graphical representations allows data to be observed from several perspectives and offers richer information than isolated representations. In this paper we propose an underlying model for an extensible and adaptable environment that allows independently developed visualization components to be gradually integrated into a user configured knowledge discovery application. Because a major requirement when using multiple visual techniques is the ability to link amongst them, so that user actions executed on a representation propagate to others if desired, the model also allows runtime configuration of coordinated user actions over different visual representations. We illustrate how this environment is being used to assist data exploration and organization in a climate classification problem.

Automatic segmentation of the secondary austenite-phase island precipitates in a superduplex stainless steel weld metal

Duplex and superduplex stainless steels are class of materials of a high importance for engineering purposes, since they have good mechanical properties combination and also are very resistant to corrosion. It is known as well that the chemical composition of such steels is very important to maintain some desired properties. In the past years, some works have reported that γ 2 precipitation improves the toughness of such steels, and its quantification may reveals some important information about steel quality. Thus, we propose in this work the automatic segmentation of γ 2 precipitation using two pattern recognition techniques: Optimum-Path Forest (OPF) and a Bayesian classifier. To the best of our knowledge, this if the first time that machine learning techniques are applied into this area. The experimental results showed that both techniques achieved similar and good recognition rates. © 2012 Taylor & Francis Group.

Further investigation of the initial fission-track length and geometry factor in apatite fission-track thermochronology

The external detector method (EDM) is a widely used technique in fission track thermochronology (FTT) in which two different minerals are concomitantly employed: spontaneous tracks are observed in apatite and induced ones in the muscovite external detector. They show intrinsic differences in detection and etching properties that should be taken into account. In this work, new geometry factor values, g, in apatite, were obtained by directly measuring the ρed/ρis ratios and independently determined [GQR]ed/is values through the measurement of projected lengths. Five mounts, two of which were large area prismatic sections and three samples composed of random-orientation pieces have been used to determine the g-values. A side effect of applying EDM is that the value of the initial confined induced fission track, L0, is not measured in routine analyses. The L 0-value is an important parameter to quantify with good confidence the degree of annealing of the spontaneous fission tracks in unknown-age samples, and is essential for accurate thermal history modeling. The impact of using arbitrary L0-values on the inference of sample thermal history is investigated and discussed. The measurement of the L0-value for each sample to be dated using an extra irradiated apatite mount is proposed. This extra mount can be also used for determining the g value as an extension of the ρed/ρis ratio method. Eight apatite samples from crystalline basement, with grains at random orientation, were used to determine the g-values. The results found are statistically in agreement with the values found for apatite samples (from Durango, Mexico) measured in prismatic section and also measured at random orientation. There was no observable variation in efficiency regarding crystal orientation, showing that it is relatively safe using non-prismatic grains, especially in samples with paucity of grains, as it is the case of most basin samples. Implications for the ζ-calibration and for the calibration of the direct (spectrometer-based) fission-track dating are also discussed.

O texto e o traço: reflexões sobre as crônicas ilustradas de Arnaldo Jabor

Pós-graduação em Comunicação - FAAC

Análise comportamental das instruções e perguntas de uma professora universitária em sala de aula de uma turma de matemática

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)