950 resultados para cross validation
Resumo:
El objetivo principal de esta tesis doctoral es profundizar en el análisis y diseño de un sistema inteligente para la predicción y control del acabado superficial en un proceso de fresado a alta velocidad, basado fundamentalmente en clasificadores Bayesianos, con el prop´osito de desarrollar una metodolog´ıa que facilite el diseño de este tipo de sistemas. El sistema, cuyo propósito es posibilitar la predicción y control de la rugosidad superficial, se compone de un modelo aprendido a partir de datos experimentales con redes Bayesianas, que ayudar´a a comprender los procesos dinámicos involucrados en el mecanizado y las interacciones entre las variables relevantes. Dado que las redes neuronales artificiales son modelos ampliamente utilizados en procesos de corte de materiales, también se incluye un modelo para fresado usándolas, donde se introdujo la geometría y la dureza del material como variables novedosas hasta ahora no estudiadas en este contexto. Por lo tanto, una importante contribución en esta tesis son estos dos modelos para la predicción de la rugosidad superficial, que se comparan con respecto a diferentes aspectos: la influencia de las nuevas variables, los indicadores de evaluación del desempeño, interpretabilidad. Uno de los principales problemas en la modelización con clasificadores Bayesianos es la comprensión de las enormes tablas de probabilidad a posteriori producidas. Introducimos un m´etodo de explicación que genera un conjunto de reglas obtenidas de árboles de decisión. Estos árboles son inducidos a partir de un conjunto de datos simulados generados de las probabilidades a posteriori de la variable clase, calculadas con la red Bayesiana aprendida a partir de un conjunto de datos de entrenamiento. Por último, contribuimos en el campo multiobjetivo en el caso de que algunos de los objetivos no se puedan cuantificar en números reales, sino como funciones en intervalo de valores. Esto ocurre a menudo en aplicaciones de aprendizaje automático, especialmente las basadas en clasificación supervisada. En concreto, se extienden las ideas de dominancia y frontera de Pareto a esta situación. Su aplicación a los estudios de predicción de la rugosidad superficial en el caso de maximizar al mismo tiempo la sensibilidad y la especificidad del clasificador inducido de la red Bayesiana, y no solo maximizar la tasa de clasificación correcta. Los intervalos de estos dos objetivos provienen de un m´etodo de estimación honesta de ambos objetivos, como e.g. validación cruzada en k rodajas o bootstrap.---ABSTRACT---The main objective of this PhD Thesis is to go more deeply into the analysis and design of an intelligent system for surface roughness prediction and control in the end-milling machining process, based fundamentally on Bayesian network classifiers, with the aim of developing a methodology that makes easier the design of this type of systems. The system, whose purpose is to make possible the surface roughness prediction and control, consists of a model learnt from experimental data with the aid of Bayesian networks, that will help to understand the dynamic processes involved in the machining and the interactions among the relevant variables. Since artificial neural networks are models widely used in material cutting proceses, we include also an end-milling model using them, where the geometry and hardness of the piecework are introduced as novel variables not studied so far within this context. Thus, an important contribution in this thesis is these two models for surface roughness prediction, that are then compared with respecto to different aspects: influence of the new variables, performance evaluation metrics, interpretability. One of the main problems with Bayesian classifier-based modelling is the understanding of the enormous posterior probabilitiy tables produced. We introduce an explanation method that generates a set of rules obtained from decision trees. Such trees are induced from a simulated data set generated from the posterior probabilities of the class variable, calculated with the Bayesian network learned from a training data set. Finally, we contribute in the multi-objective field in the case that some of the objectives cannot be quantified as real numbers but as interval-valued functions. This often occurs in machine learning applications, especially those based on supervised classification. Specifically, the dominance and Pareto front ideas are extended to this setting. Its application to the surface roughness prediction studies the case of maximizing simultaneously the sensitivity and specificity of the induced Bayesian network classifier, rather than only maximizing the correct classification rate. Intervals in these two objectives come from a honest estimation method of both objectives, like e.g. k-fold cross-validation or bootstrap.
Resumo:
Esta Tesis doctoral fue desarrollada para estudiar las emisiones de amoniaco (NH3) y metano (CH4) en purines de cerdos, y los efectos ocasionados por cambios en la formulación de la dieta. Con este propósito, fueron llevados a cabo tres estudios. El experimento 1 fue realizado con el objetivo de analizar los factores de variación de la composición de purines y establecer ecuaciones de predicción para emisiones potenciales de NH3 y CH4. Fueron recogidas setenta y nueve muestras de piensos y purines durante dos estaciones del año (verano y invierno) de granjas comerciales situadas en dos regiones de España (Centro y Mediterráneo). Se muestrearon granjas de gestación, maternidad, lactación y cebo. Se determinó la composición de piensos y purines, y la emisión potencial de NH3 y CH4. El contenido de nutrientes de los piensos fue usado como covariable en el análisis. La espectroscopia de reflectancia del infrarrojo cercano (NIRS) se evaluó como herramienta de predicción de la composición y potencial emisión de gases del purín. Se encontró una amplia variabilidad en la composición de piensos y purines. Las granjas del Mediterráneo tenían mayor pH (P<0,001) y concentración de cenizas (P =0,02) en el purín que las del Centro. El tipo de granja también afectó al contenido de extracto etéreo (EE) del purín (P =0,02), observando los valores más elevados en las instalaciones de animales jóvenes. Los resultados sugieren un efecto tampón de la fibra de la dieta en el pH del purín y una relación directa (P<0,05) con el contenido de fibra fecal. El contenido de proteína del pienso no afectó al contenido de nitrógeno del purín, pero disminuyó (P=0,003) la concentración de sólidos totales (ST) y de sólidos volátiles (SV). Se obtuvieron modelos de predicción de la emisión potencial de NH3 (R2=0,89) y CH4 (R2=0,61) partir de la composición del purín. Los espectros NIRS mostraron una buena precisión para la estimación de la mayor parte de los constituyentes, con coeficientes de determinación de validación cruzada (R2cv) superiores a 0,90, así como para la predicción del potencial de emisiones de NH3 y CH4 (R2cv=0,84 y 0,68, respectivamente). El experimento 2 fue realizado para investigar los efectos del nivel de inclusión de dos fuentes de sub-productos fibrosos: pulpa de naranja (PN) y pulpa de algarroba (PA), en dietas iso-fibrosas de cerdos de cebo, sobre la composición del purín y las emisiones potenciales de NH3 y CH4. Treinta cerdos (85,4±12,3 kg) fueron alimentados con cinco dietas iso-nutritivas: control comercial trigo/cebada (C) y cuatro dietas experimentales incluyendo las dos fuentes de sub-productos a dos niveles (75 y 150 g/kg) en una estructura 2 × 2 factorial. Después de 14 días de periodo de adaptación, heces y orina fueron recogidas separadamente durante 7 días para medir la digestibilidad de los nutrientes y el nitrógeno (N) excretado (6 réplicas por dieta) en cerdos alojados individualmente en jaulas metabólicas. Las emisiones de NH3 y CH4 fueron medidas después de la recogida de los purínes durante 11 y 100 días respectivamente. La fuente y el nivel de subproductos fibrosos afectó a la eficiencia digestiva de diferentes formas, ya que los coeficientes de digestibilidad total aparente (CDTA) para la materia seca (MS), materia orgánica (MO), fracciones fibrosas y energía bruta (EB) aumentaron con la PN pero disminuyeron con la inclusión de PA (P<0,05). El CDTA de proteína bruta (PB) disminuyó con la inclusión de las dos fuentes de fibra, siendo más bajo al mayor nivel de inclusión. La concentración fecal de fracciones fibrosas aumentó (P<0,05) con el nivel de inclusión de PA pero disminuyó con el de PN (P<0,01). El nivel más alto de las dos fuentes de fibra en el pienso aumentó (P<0,02) el contenido de PB fecal pero disminuyó el contenido de N de la orina (de 205 para 168 g/kg MS, P<0,05) en todas las dietas suplementadas comparadas con la dieta C. Adicionalmente, las proporciones de nitrógeno indigerido, nitrógeno soluble en agua, nitrógeno bacteriano y endógeno excretado en heces no fueron afectados por los tratamientos. Las características iniciales del purín no difirieron entre las diferentes fuentes y niveles de fibra, excepto para el pH que disminuyó con la inclusión de altos niveles de sub-productos. La emisión de NH3 por kg de purín fue más baja en todas las dietas suplementadas con fibras que en la dieta C (2,44 vs.1,81g de promedio, P<0,05). Además, purines de dietas suplementadas con alto nivel de sub-productos tendieron (P<0,06) a emitir menos NH3 por kg de nitrógeno total y mostraron un potencial más bajo para emitir CH4, independientemente de la fuente de fibra. El experimento 3 investigó los efectos de la fuente de proteína en dietas prácticas. Tres piensos experimentales fueron diseñados para sustituir una mescla de harina y cascarilla de soja (SOJ) por harina de girasol (GIR) o por DDGS del trigo (DDGST). La proporción de otros ingredientes fue modificada para mantener los contenidos de nutrientes similares a través de las dietas. El cambio en la fuente de proteína dio lugar a diferencias en el contenido de fibra neutro detergente ligada a proteína bruta (FNDPB), fibra soluble (FS) y lignina ácido detergente (LAD) en la dieta. Veinticuatro cerdos (ocho por dieta), con 52,3 o 60,8 kg en la primera y segunda tanda respectivamente, fueron alojados individualmente en jaulas metabólicas. Durante un periodo de 7 días fue determinado el balance de MS, el CDTA de los nutrientes y la composición de heces y orina. Se realizó el mismo procedimiento del experimento 2 para medir las emisiones de NH3 y CH4 de los purines de cada animal. Ni la ingestión de MS ni el CDTA de la MS o de la energía fueron diferentes entre las dietas experimentales, pero el tipo de pienso afectó (P<0.001) la digestibilidad de la PB, que fue mayor para GIR (0,846) que para SOJ (0,775), mientras que la dieta DDGST mostró un valor intermedio (0,794). La concentración fecal de PB fue por tanto influenciada (P<0,001) por el tratamiento, observándose la menor concentración de PB en la dieta GIR y la mayor en la dieta SOJ. La proporción de N excretado en orina o heces disminuyó de 1,63 en la dieta GIR hasta 0,650 en la dieta SOJ, como consecuencia de perdidas más bajas en orina y más altas en heces, con todas las fracciones de nitrógeno fecales creciendo en paralelo a la excreción total. Este resultado fue paralelo a una disminución de la emisión potencial de NH3 (g/kg purín) en la dieta SOJ con respecto a la dieta GIR (desde 1,82 a 1,12, P<0,05), dando valores intermedios (1,58) para los purines de la dieta DDGST. Por otro lado, el CDTA de la FS y de la fibra neutro detergente (FND) fueron afectados (P<0,001 y 0,002, respectivamente) por el tipo de dieta, siendo más bajas en la dieta GIR que en la dieta SOJ; además, se observó un contenido más alto de FND (491 vs. 361g/kg) en la MS fecal para la dieta GIR que en la dieta SOJ, presentando la dieta DDGST valores intermedios. El grado de lignificación de la FND (FAD/FND x 100) de las heces disminuyó en el orden GIR>DDGST>SOJ (desde 0,171 hasta 0,109 y 0,086, respectivamente) en paralelo a la disminución del potencial de emisión de CH4 por g de SV del purín (desde 301 a 269 y 256 mL, respectivamente). Todos los purines obtenidos en estos tres experimentos y Antezana et al. (2015) fueron usados para desarrollar nuevas calibraciones con la tecnología NIRS, para predecir la composición del purín y el potencial de las emisiones de gases. Se observó una buena precisión (R2cv superior a 0,92) de las calibraciones cuando muestras de los ensayos controlados (2, 3 y Antezana et al., 2015) fueron añadidas, aumentando el rango de variación. Una menor exactitud fue observada para TAN y emisiones de NH3 y CH4, lo que podría explicarse por una menor homogeneidad en la distribución de las muestras cuando se amplía el rango de variación del estudio. ABSTRACT This PhD thesis was developed to study the emissions of ammonia (NH3) and methane (CH4) from pig slurry and the effects caused by changes on diet formulation. For these proposes three studies were conducted. Experiment 1 aimed to analyse several factors of variation of slurry composition and to establish prediction equations for potential CH4 and NH3 emissions. Seventy-nine feed and slurry samples were collected at two seasons (summer and winter) from commercial pig farms sited at two Spanish regions (Centre and Mediterranean). Nursery, growing-fattening, gestating and lactating facilities were sampled. Feed and slurry composition were determined, and potential CH4 and NH3 emissions measured. Feed nutrient contents were used as covariates in the analysis. Near infrared reflectance spectroscopy (NIRS) was evaluated as a predicting tool for slurry composition and potential gaseous emissions. A wide variability was found both in feed and slurry composition. Mediterranean farms had a higher pH (P<0.001) and ash (P=0.02) concentration than those located at the centre of Spain. Also, type of farm affected ether extract (EE) content of the slurry (P=0.02), with highest values obtained for the youngest animal facilities. Results suggested a buffer effect of dietary fibre on slurry pH and a direct relationship (P<0.05) with fibre constituents of manure. Dietary protein content did not affect slurry nitrogen content (N) but decreased (P=0.003) in total solid (TS) and volatile solids (VS) concentration. Prediction models of potential NH3 emissions (R2=0.89) and biochemical CH4 potential (B0) (R2=0.61) were obtained from slurry composition. Predictions from NIRS showed a high accuracy for most slurry constituents with coefficient of determination of cross validation (R2cv) above 0.90 and a similar accuracy of prediction of potential NH3 and CH4 emissions (R2cv=0.84 and 0.68, respectively) thus models based on slurry composition from commercial farms. Experiment 2 was conducted to investigate the effects of increasing the level of two sources of fibrous by-products, orange pulp (OP) and carob meal (CM), in iso-fibrous diets for growing-finishing pig, slurry composition and potential NH3 and CH4 emissions. Thirty pigs (85.4±12.3 kg) were fed five iso-nutritive diets: a commercial control wheat/barley (C) and four experimental diets including two sources of fibrous by-products OP and CM and two dietary levels (75 and 150 g/kg) in a 2 × 2 factorial arrangement. After a 14-day adaptation period, faeces and urine were collected separately for 7 days to measure nutrient digestibility and the excretory patterns of N from pigs (6 replicates per diet) housed individually in metabolic pens. For each animal, the derived NH3 and CH4 emissions were measured in samples of slurry over an 11 and 100-day storage periods, respectively. Source and level of the fibrous by-products affected digestion efficiency in a different way as the coefficients of total tract apparent digestibility (CTTAD) for dry matter (DM), organic matter (OM), fibre fractions and gross energy (GE) increased with OP but decreased with CM (P<0.05). Crude protein CTTAD decreased with the inclusion of both sources of fibre, being lower at the highest dietary level. Faecal concentration of fibre fractions increased (P<0.05) with the level of inclusion of CM but decreased with that of OP (P<0.01). High dietary level for both sources of fibre increased (P<0.02) CP faecal content but urine N content decreased (from 205 to 168 g/kg DM, P<0.05) in all the fibre-supplemented compared to C diet. Additionally, the proportions of undigested dietary, water soluble, and bacterial and endogenous debris of faecal N excretion were not affected by treatments. The initial slurry characteristics did not differ among different fibre sources and dietary levels, except pH, which decreased at the highest by-product inclusion levels. Ammonia emission per kg of slurry was lower in all the fibre-supplemented diets than in C diet (2.44 vs. 1.81g as average, P<0.05). Additionally, slurries from the highest dietary level of by-products tended (P<0.06) to emit less NH3 per kg of initial total Kjeldahl nitrogen (TKN) and showed a lower biochemical CH4 potential , independently of the fibre source. Experiment 3 investigated the effects of protein source in practical diets. Three experimental feeds were designed to substitute a mixture of soybean meal and soybean hulls (SB diet) with sunflower meal (SFM) or wheat DDGS (WDDGS). The proportion of other ingredients was also modified in order to maintain similar nutrient contents across diets. Changes in protein source led to differences in dietary content of neutral detergent insoluble crude protein (NDICP), soluble fibre (SF) and acid detergent lignin (ADL). Twenty-four pigs (eight per diet), weighing 52.3 or 60.8 kg at the first and second batch respectively, were housed individually in metabolic pens to determine during a 7-day period DM balance, CTTAD of nutrients, and faecal and urine composition. Representative slurry samples from each animal were used to measure NH3 and CH4 emissions over an 11 and or 100-day storage period, respectively. Neither DM intake, nor DM or energy CTTAD differed among experimental diets, but type of feed affected (P<0.001) CP digestibility, which was highest for SFM (0.846) than for SB (0.775) diet, with WDDGS-based diet giving an intermediate value (0.794). Faecal DM composition was influenced (P<0.001) accordingly, with the lowest CP concentration found for diet SFM and the highest for SB. The ratio of N excreted in urine or faeces decreased from SFM (1.63) to SB diet (0.650), as a consequence of both lower urine and higher faecal losses, with all the faecal N fractions increasing in parallel to total excretion. This result was parallel to a decrease of potential NH3 emission (g/kg slurry) in diet SB with respect to diet SFM (from 1.82 to 1.12, P<0.05), giving slurry from WDDGS-based diet an intermediate value (1.58). Otherwise, SF and insoluble neutral detergent fibre (NDF) CTTAD were affected (P<0.001 and P=0.002, respectively) by type of diet, being lower for SFM than in SB-diet; besides, a higher content of NDF (491 vs. 361 g/kg) in faecal DM was observed for SFM with respect to SB based diet, with WDDGS diet being intermediate. Degree of lignification of NDF (ADL/NDF x 100) of faeces decreased in the order SFM>WDDGS>SB (from 0.171 to 0.109 and 0.086, respectively) in parallel to a decrease of biochemical CH4 potential per g of VS of slurry (from 301 to 269 and 256 ml, respectively). All slurry samples obtained from these three experiments and Antezana et al. (2015) were used to develop new calibrations with NIRS technology, to predict the slurry composition and potential gaseous emissions of samples with greater variability in comparison to experiment 1. Better accuracy (R2cv above 0.92) was observed for calibrations when samples from controlled trials experiments (2, 3 and Antezana et al., 2015) were included, increasing the range of variation. A lower accuracy was observed for TAN, NH3 and CH4 gaseous emissions, which might be explained by the less homogeneous distribution with a wider range of data.
Resumo:
En los últimos años han surgido nuevos campos de las tecnologías de la información que exploran el tratamiento de la gran cantidad de datos digitales existentes y cómo transformarlos en conocimiento explícito. Las técnicas de Procesamiento del Lenguaje Natural (NLP) son capaces de extraer información de los textos digitales presentados en forma narrativa. Además, las técnicas de machine learning clasifican instancias o ejemplos en función de sus atributos, en distintas categorías, aprendiendo de otros previamente clasificados. Los textos clínicos son una gran fuente de información no estructurada; en consecuencia, información no explotada en su totalidad. Algunos términos usados en textos clínicos se encuentran en una situación de afirmación, negación, hipótesis o histórica. La detección de esta situación es necesaria para la estructuración de información, pero a su vez tiene una gran complejidad. Extrayendo características lingüísticas de los elementos, o tokens, de los textos mediante NLP; transformando estos tokens en instancias y las características en atributos, podemos mediante técnicas de machine learning clasificarlos con el objetivo de detectar si se encuentran afirmados, negados, hipotéticos o históricos. La selección de los atributos que cada token debe tener para su clasificación, así como la selección del algoritmo de machine learning utilizado son elementos cruciales para la clasificación. Son, de hecho, los elementos que componen el modelo de clasificación. Consecuentemente, este trabajo aborda el proceso de extracción de características, selección de atributos y selección del algoritmo de machine learning para la detección de la negación en textos clínicos en español. Se expone un modelo para la clasificación que, mediante el algoritmo J48 y 35 atributos obtenidos de características lingüísticas (morfológicas y sintácticas) y disparadores de negación, detecta si un token está negado en 465 frases provenientes de textos clínicos con un F-Score del 73%, una exhaustividad del 66% y una precisión del 81% con una validación cruzada de 10 iteraciones. ---ABSTRACT--- New information technologies have emerged in the recent years which explore the processing of the huge amount of existing digital data and its transformation into knowledge. Natural Language Processing (NLP) techniques are able to extract certain features from digital texts. Additionally, through machine learning techniques it is feasible to classify instances according to different categories, learning from others previously classified. Clinical texts contain great amount of unstructured data, therefore information not fully exploited. Some terms (tokens) in clinical texts appear in different situations such as affirmed, negated, hypothetic or historic. Detecting this situation is necessary for the structuring of this data, however not simple. It is possible to detect whether if a token is negated, affirmed, hypothetic or historic by extracting its linguistic features by NLP; transforming these tokens into instances, the features into attributes, and classifying these instances through machine learning techniques. Selecting the attributes each instance must have, and choosing the machine learning algorithm are crucial issues for the classification. In fact, these elements set the classification model. Consequently, this work approaches the features retrieval as well as the attributes and algorithm selection process used by machine learning techniques for the detection of negation in clinical texts in Spanish. We present a classification model which, through J48 algorithm and 35 attributes from linguistic features (morphologic and syntactic) and negation triggers, detects whether if a token is negated in 465 sentences from historical records, with a result of 73% FScore, 66% recall and 81% precision using a 10-fold cross-validation.
Resumo:
We present a set of new volume scaling relationships specific to Svalbard glaciers, derived from a sample of 60 volume–area pairs. Glacier volumes are computed from ground-penetrating radar (GPR)-retrieved ice thickness measurements, which have been compiled from different sources for this study. The most precise scaling models, in terms of lowest cross-validation errors, are obtained using a multivariate approach where, in addition to glacier area, glacier length and elevation range are also used as predictors. Using this multivariate scaling approach, together with the Randolph Glacier Inventory V3.2 for Svalbard and Jan Mayen, we obtain a regional volume estimate of 6700 ± 835 km3, or 17 ± 2 mm of sea-level equivalent (SLE). This result lies in the mid- to low range of recently published estimates, which show values as varied as 13 and 24 mm SLE. We assess the sensitivity of the scaling exponents to glacier characteristics such as size, aspect ratio and average slope, and find that the volume of steep-slope and cirque-type glaciers is not very sensitive to changes in glacier area.
Resumo:
We present a method for predicting protein folding class based on global protein chain description and a voting process. Selection of the best descriptors was achieved by a computer-simulated neural network trained on a data base consisting of 83 folding classes. Protein-chain descriptors include overall composition, transition, and distribution of amino acid attributes, such as relative hydrophobicity, predicted secondary structure, and predicted solvent exposure. Cross-validation testing was performed on 15 of the largest classes. The test shows that proteins were assigned to the correct class (correct positive prediction) with an average accuracy of 71.7%, whereas the inverse prediction of proteins as not belonging to a particular class (correct negative prediction) was 90-95% accurate. When tested on 254 structures used in this study, the top two predictions contained the correct class in 91% of the cases.
Resumo:
As análises biplot que utilizam os modelos de efeitos principais aditivos com inter- ação multiplicativa (AMMI) requerem matrizes de dados completas, mas, frequentemente os ensaios multiambientais apresentam dados faltantes. Nesta tese são propostas novas metodologias de imputação simples e múltipla que podem ser usadas para analisar da- dos desbalanceados em experimentos com interação genótipo por ambiente (G×E). A primeira, é uma nova extensão do método de validação cruzada por autovetor (Bro et al, 2008). A segunda, corresponde a um novo algoritmo não-paramétrico obtido por meio de modificações no método de imputação simples desenvolvido por Yan (2013). Também é incluído um estudo que considera sistemas de imputação recentemente relatados na literatura e os compara com o procedimento clássico recomendado para imputação em ensaios (G×E), ou seja, a combinação do algoritmo de Esperança-Maximização com os modelos AMMI ou EM-AMMI. Por último, são fornecidas generalizações da imputação simples descrita por Arciniegas-Alarcón et al. (2010) que mistura regressão com aproximação de posto inferior de uma matriz. Todas as metodologias têm como base a decomposição por valores singulares (DVS), portanto, são livres de pressuposições distribucionais ou estruturais. Para determinar o desempenho dos novos esquemas de imputação foram realizadas simulações baseadas em conjuntos de dados reais de diferentes espécies, com valores re- tirados aleatoriamente em diferentes porcentagens e a qualidade das imputações avaliada com distintas estatísticas. Concluiu-se que a DVS constitui uma ferramenta útil e flexível na construção de técnicas eficientes que contornem o problema de perda de informação em matrizes experimentais.
Resumo:
As doenças tropicais negligenciadas (DTNs) causam um imenso sofrimento para a pessoa acometida e em muitos casos podem levar o indivíduo a morte. Elas representam um obstáculo devastador para a saúde e continuam a ser um sério impedimento para a redução da pobreza e desenvolvimento socioeconômico. Das 17 doenças desse grupo, a leishmaniose, incluindo a leishmaniose cutânea, tem grande destaque devido sua alta incidência, os gastos para o tratamento e as complicações geradas em processos de coinfecção. Ainda mais agravante, os investimentos direcionados ao controle, combate e principalmente a inovação em novos produtos é ainda muito limitado. Atualmente, a academia tem um importante papel na luta contra essas doenças através da busca de novos alvos terapêuticos e também de novas moléculas com potencial terapêutico. É nesse contexto que esse projeto teve como meta a implantação de uma plataforma para a identificação de moléculas com atividade leishmanicida. Como alvo terapêutico, optamos pela utilização da enzima diidroorotato desidrogenase de Leishmania Viannia braziliensis (LbDHODH), enzima de extrema importância na síntese de novo de nucleotídeos de pirimidina, cuja principal função é converter o diidroorotato em orotato. Esta enzima foi clonada, expressa e purificada com sucesso em nosso laboratório. Os estudos permitiram que a enzima fosse caracterizada cineticamente e estruturalmente via cristalografia de raios- X. Os primeiros ensaios inibitórios foram realizados com o orotato, produto da catálise e inibidor natural da enzima. O potencial inibitório do orotato foi mensurado através da estimativa do IC50 e a interação proteína-ligante foi caracterizada através de estudos cristalográficos. Estratégias in silico e in vitro foram utilizadas na busca de ligantes, através das quais foram identificados inibidores para a enzima LbDHODH. Ensaios de validação cruzada, utilizando a enzima homóloga humana, permitiram identificar os ligantes com maior índice de seletividade que tiveram seu potencial leishmanicida avaliado in vitro contra as formas promastigota e amastigota de Leishmania braziliensis. A realização do presente projeto permitiu a identificação de uma classe de ligantes que apresentam atividade seletiva contra LbDHODH e que será utilizada no planejamento de futuras gerações de moléculas com atividade terapêutica para o tratamento da leishmaniose. Além disso, a plataforma de ensaios otimizada permitirá a avaliação de novos grupos de moléculas como uma importante estratégia na busca por novos tratamentos contra a leishmaniose
Resumo:
A condutividade hidráulica (K) é um dos parâmetros controladores da magnitude da velocidade da água subterrânea, e consequentemente, é um dos mais importantes parâmetros que afetam o fluxo subterrâneo e o transporte de solutos, sendo de suma importância o conhecimento da distribuição de K. Esse trabalho visa estimar valores de condutividade hidráulica em duas áreas distintas, uma no Sistema Aquífero Guarani (SAG) e outra no Sistema Aquífero Bauru (SAB) por meio de três técnicas geoestatísticas: krigagem ordinária, cokrigagem e simulação condicional por bandas rotativas. Para aumentar a base de dados de valores de K, há um tratamento estatístico dos dados conhecidos. O método de interpolação matemática (krigagem ordinária) e o estocástico (simulação condicional por bandas rotativas) são aplicados para estimar os valores de K diretamente, enquanto que os métodos de krigagem ordinária combinada com regressão linear e cokrigagem permitem incorporar valores de capacidade específica (Q/s) como variável secundária. Adicionalmente, a cada método geoestatístico foi aplicada a técnica de desagrupamento por célula para comparar a sua capacidade de melhorar a performance dos métodos, o que pode ser avaliado por meio da validação cruzada. Os resultados dessas abordagens geoestatísticas indicam que os métodos de simulação condicional por bandas rotativas com a técnica de desagrupamento e de krigagem ordinária combinada com regressão linear sem a técnica de desagrupamento são os mais adequados para as áreas do SAG (rho=0.55) e do SAB (rho=0.44), respectivamente. O tratamento estatístico e a técnica de desagrupamento usados nesse trabalho revelaram-se úteis ferramentas auxiliares para os métodos geoestatísticos.
Resumo:
This paper presents the first version of EmotiBlog, an annotation scheme for emotions in non-traditional textual genres such as blogs or forums. We collected a corpus composed by blog posts in three languages: English, Spanish and Italian and about three topics of interest. Subsequently, we annotated our collection and carried out the inter-annotator agreement and a ten-fold cross-validation evaluation, obtaining promising results. The main aim of this research is to provide a finer-grained annotation scheme and annotated data that are essential to perform evaluation focused on checking the quality of the created resources.
Resumo:
Purpose. To assess in a sample of normal, keratoconic, and keratoconus (KC) suspect eyes the performance of a set of new topographic indices computed directly from the digitized images of the Placido rings. Methods. This comparative study was composed of a total of 124 eyes of 106 patients from the ophthalmic clinics Vissum Alicante and Vissum Almería (Spain) divided into three groups: control group (50 eyes), KC group (50 eyes), and KC suspect group (24 eyes). In all cases, a comprehensive examination was performed, including the corneal topography with a Placidobased CSO topography system. Clinical outcomes were compared among groups, along with the discriminating performance of the proposed irregularity indices. Results. Significant differences at level 0.05 were found on the values of the indices among groups by means of Mann-Whitney-Wilcoxon nonparametric test and Fisher exact test. Additional statistical methods, such as receiver operating characteristic analysis and K-fold cross validation, confirmed the capability of the indices to discriminate between the three groups. Conclusions. Direct analysis of the digitized images of the Placido mires projected on the cornea is a valid and effective tool for detection of corneal irregularities. Although based only on the data from the anterior surface of the cornea, the new indices performed well even when applied to the KC suspect eyes. They have the advantage of simplicity of calculation combined with high sensitivity in corneal irregularity detection and thus can be used as supplementary criteria for diagnosing and grading KC that can be added to the current keratometric classifications.
Resumo:
Background: The assessment of attitudes toward school with the objective of identifying adolescents who may be at risk of underachievement has become an important area of research in educational psychology, although few specific tools for their evaluation have been designed to date. One of the instruments available is the School Attitude Assessment Survey-Revised (SAAS-R). Method: The objective of the current research is to test the construct validity and to analyze the psychometric properties of the Spanish version of the SAAS-R. Data were collected from 1,398 students attending different high schools. Students completed the SAAS-R along with measures of the g factor, and academic achievement was obtained from school records. Results: Confirmatory factor analysis, multivariate analysis of variance and analysis of variance tests supported the validity evidence. Conclusions: The results indicate that the Spanish version of the SAAS-R is a useful measure that contributes to identification of underachieving students. Lastly, the results obtained and their implications for education are discussed.
Resumo:
We examined the psychometric properties of the School Attitude Assessment Survey–Revised in a Spanish population (n = 1,398). Confirmatory factor analysis procedures supported the instrument’s five-factor structure. The results of discriminant analysis demonstrated the predictive power of the School Attitude Assessment Survey–Revised scales as regards academic performance. Implications for education and assessment are discussed.
Resumo:
Numerical modelling methodologies are important by their application to engineering and scientific problems, because there are processes where analytical mathematical expressions cannot be obtained to model them. When the only available information is a set of experimental values for the variables that determine the state of the system, the modelling problem is equivalent to determining the hyper-surface that best fits the data. This paper presents a methodology based on the Galerkin formulation of the finite elements method to obtain representations of relationships that are defined a priori, between a set of variables: y = z(x1, x2,...., xd). These representations are generated from the values of the variables in the experimental data. The approximation, piecewise, is an element of a Sobolev space and has derivatives defined in a general sense into this space. The using of this approach results in the need of inverting a linear system with a structure that allows a fast solver algorithm. The algorithm can be used in a variety of fields, being a multidisciplinary tool. The validity of the methodology is studied considering two real applications: a problem in hydrodynamics and a problem of engineering related to fluids, heat and transport in an energy generation plant. Also a test of the predictive capacity of the methodology is performed using a cross-validation method.
Resumo:
Tese de mestrado, Bioinformática e Biologia Computacional (Bioinformática), Universidade de Lisboa, Faculdade de Ciências, 2016
Resumo:
Individual analysis of functional Magnetic Resonance Imaging (fMRI) scans requires user-adjustment of the statistical threshold in order to maximize true functional activity and eliminate false positives. In this study, we propose a novel technique that uses radiomic texture analysis (TA) features associated with heterogeneity to predict areas of true functional activity. Scans of 15 right-handed healthy volunteers were analyzed using SPM8. The resulting functional maps were thresholded to optimize visualization of language areas, resulting in 116 regions of interests (ROIs). A board-certified neuroradiologist classified different ROIs into Expected (E) and Non-Expected (NE) based on their anatomical locations. TA was performed using the mean Echo-Planner Imaging (EPI) volume, and 20 rotation-invariant texture features were obtained for each ROI. Using forward stepwise logistic regression, we built a predictive model that discriminated between E and NE areas of functional activity, with a cross-validation AUC and success rate of 79.84% and 80.19% respectively (specificity/sensitivity of 78.34%/82.61%). This study found that radiomic TA of fMRI scans may allow for determination of areas of true functional activity, and thus eliminate clinician bias.
