Made available in DSpace on 2017-07-10T19:24:46Z (GMT). No. of bitstreams: 1 Marcio Paulo de Oliveira.pdf: 1960743 bytes, checksum: 7438fe00d388d47b01b27d6cfdf2e229 (MD5) Previous issue date: 2010-02-04 / The knowledge about soil and plant attributes is important for the improvement of agricultural management. Intense tillage activities may induce not only alterations in the soil attributes but also decrease in productivity. Studies directed to the soil and plant spatial variability identification and the relations amid these variables are tools for agriculture, with the potential to increase productivity. The data set for this study was sampled in a Rhodic Acrudox soil, at a farmland that has been being cultivated for over five years under no-tillage system, with soybean and wheat in crop succession. At 252 m long transect, 84 points were demarcated, with 3 m of spacing between each of them. The relations between soybean productivity and soil water content, micro, macro and total porosity, soil density and soil resistance to penetration at 0,0-0,10 m and 0,10-0,20 m deep layers, were evaluated, as well as the respective variabilities. The relations between soybean productivity and soil attributes were determined using simple and cross correlations, followed by the state space models determinations, compared to linear and multiple regression models. The results have shown that the soybean productivity and soil mechanical resistance variables presented not only autocorrelation structure but also crosscorrelation structure. The state space models, relating to the soybean productivity at a point i, with the same attribute at point i-1, at the two layers, were more efficient than the equivalent models in simple and multiple regression. With geoestatistics, the spatial dependence structure was determined with envelopes and models for the semivariograms, allowing identification and classification of the spatial dependence for the variables under study. The thematic maps were obtained with simple kriging and indicated the soil attributes behavior, related to the soybean productivity. / O conhecimento do comportamento dos atributos do solo e da planta é importante para a melhoria das práticas agrícolas. A intensa atividade de cultivo pode provocar modificações dos atributos do solo e reduzir a produtividade de uma cultura em determinada região. Os estudos que visam identificar a variabilidade espacial dos atributos do solo e da planta e a relação entre esses atributos surgem como um recurso para a agricultura, podendo ser utilizados para realização de um manejo adequado dos recursos disponíveis, ampliando a produtividade e preservando o meioambiente. Os dados para a realização deste estudo foram obtidos em um Latossolo Vermelho distroférrico, em uma área cultivada há mais de cinco anos com alternância entre as culturas de soja e trigo, com o sistema de plantio direto. Em uma transeção de 252 m de comprimento foram demarcados 84 elementos amostrais, espaçados de 3 m entre si. As relações da produtividade da soja com os seguintes atributos físicos e hídricos do solo: teor de água no solo, microporosidade, macroporosidade e porosidade total do solo, densidade do solo e resistência mecânica do solo à penetração, nas camadas 0,0-0,10 m e 0,10-0,20 m, foram avaliadas bem como a variabilidade espacial desses atributos. A relação entre a produtividade da soja e os atributos do solo foi determinada através das correlações simples e cruzada entre os elementos amostrais de cada atributo, seguida da estimação dos modelos em espaço de estados, comparados aos modelos equivalentes em regressão linear múltipla. Os resultados mostraram que as variáveis produtividade da soja e resistência do solo a penetração apresentaram estrutura de autocorrelação e de correlação cruzada entre si. Os modelos estimados em espaço de estados, relacionando a produtividade da soja em um ponto i com a produtividade da soja e resistência do solo a penetração nas duas camadas no ponto i -1 mostraram-se mais eficientes do que os modelos equivalentes estimados em regressão linear simples e múltipla. Por meio da geoestatística, a estrutura de dependência espacial foi avaliada por meio dos envelopes e modelos para os semivariogramas experimentais, permitindo identificar e classificar a dependência espacial das variáveis em estudo. Os mapas temáticos foram obtidos por meio de interpolação por krigagem ordinária e indicaram o comportamento dos atributos do solo ligadas a produtividade da soja.

Made available in DSpace on 2017-05-12T14:48:09Z (GMT). No. of bitstreams: 1 Marcio Paulo de Oliveira.pdf: 1960743 bytes, checksum: 7438fe00d388d47b01b27d6cfdf2e229 (MD5) Previous issue date: 2010-02-04 / The knowledge about soil and plant attributes is important for the improvement of agricultural management. Intense tillage activities may induce not only alterations in the soil attributes but also decrease in productivity. Studies directed to the soil and plant spatial variability identification and the relations amid these variables are tools for agriculture, with the potential to increase productivity. The data set for this study was sampled in a Rhodic Acrudox soil, at a farmland that has been being cultivated for over five years under no-tillage system, with soybean and wheat in crop succession. At 252 m long transect, 84 points were demarcated, with 3 m of spacing between each of them. The relations between soybean productivity and soil water content, micro, macro and total porosity, soil density and soil resistance to penetration at 0,0-0,10 m and 0,10-0,20 m deep layers, were evaluated, as well as the respective variabilities. The relations between soybean productivity and soil attributes were determined using simple and cross correlations, followed by the state space models determinations, compared to linear and multiple regression models. The results have shown that the soybean productivity and soil mechanical resistance variables presented not only autocorrelation structure but also crosscorrelation structure. The state space models, relating to the soybean productivity at a point i, with the same attribute at point i-1, at the two layers, were more efficient than the equivalent models in simple and multiple regression. With geoestatistics, the spatial dependence structure was determined with envelopes and models for the semivariograms, allowing identification and classification of the spatial dependence for the variables under study. The thematic maps were obtained with simple kriging and indicated the soil attributes behavior, related to the soybean productivity. / O conhecimento do comportamento dos atributos do solo e da planta é importante para a melhoria das práticas agrícolas. A intensa atividade de cultivo pode provocar modificações dos atributos do solo e reduzir a produtividade de uma cultura em determinada região. Os estudos que visam identificar a variabilidade espacial dos atributos do solo e da planta e a relação entre esses atributos surgem como um recurso para a agricultura, podendo ser utilizados para realização de um manejo adequado dos recursos disponíveis, ampliando a produtividade e preservando o meioambiente. Os dados para a realização deste estudo foram obtidos em um Latossolo Vermelho distroférrico, em uma área cultivada há mais de cinco anos com alternância entre as culturas de soja e trigo, com o sistema de plantio direto. Em uma transeção de 252 m de comprimento foram demarcados 84 elementos amostrais, espaçados de 3 m entre si. As relações da produtividade da soja com os seguintes atributos físicos e hídricos do solo: teor de água no solo, microporosidade, macroporosidade e porosidade total do solo, densidade do solo e resistência mecânica do solo à penetração, nas camadas 0,0-0,10 m e 0,10-0,20 m, foram avaliadas bem como a variabilidade espacial desses atributos. A relação entre a produtividade da soja e os atributos do solo foi determinada através das correlações simples e cruzada entre os elementos amostrais de cada atributo, seguida da estimação dos modelos em espaço de estados, comparados aos modelos equivalentes em regressão linear múltipla. Os resultados mostraram que as variáveis produtividade da soja e resistência do solo a penetração apresentaram estrutura de autocorrelação e de correlação cruzada entre si. Os modelos estimados em espaço de estados, relacionando a produtividade da soja em um ponto i com a produtividade da soja e resistência do solo a penetração nas duas camadas no ponto i -1 mostraram-se mais eficientes do que os modelos equivalentes estimados em regressão linear simples e múltipla. Por meio da geoestatística, a estrutura de dependência espacial foi avaliada por meio dos envelopes e modelos para os semivariogramas experimentais, permitindo identificar e classificar a dependência espacial das variáveis em estudo. Os mapas temáticos foram obtidos por meio de interpolação por krigagem ordinária e indicaram o comportamento dos atributos do solo ligadas a produtividade da soja.

Accurate long-term energy consumption forecasting of industrial entities is of interest to distribution companies as it can potentially help reduce their churn and offer support in decision making when hedging. This thesis work presents different methods to forecast the energy consumption for industrial entities over a long time prediction horizon of 1 year. Notably, it includes experimentations with two variants of the Recurrent Neural Networks, namely Gated Recurrent Unit (GRU) and Long-Short-Term-Memory (LSTM). Their performance is compared against traditional approaches namely Multiple Linear Regression (MLR) and Seasonal Autoregressive Integrated Moving Average (SARIMA). Further on, the investigation focuses on tailoring the Recurrent Neural Network model to improve the performance. The experiments focus on the impact of different model architectures. Secondly, it focuses on testing the effect of time-related feature selection as an additional input to the Recurrent Neural Network (RNN) networks. Specifically, it explored how traditional methods such as Exploratory Data Analysis, Autocorrelation, and Partial Autocorrelation Functions Plots can contribute to the performance of RNN model. The current work shows through an empirical study on three industrial datasets that GRU architecture is a powerful method for the long-term forecasting task which outperforms LSTM on certain scenarios. In comparison to the MLR model, the RNN achieved a reduction in the RMSE between 5% up to to 10%. The most important findings include: (i) GRU architecture outperforms LSTM on industrial energy consumption datasets when compared against a lower number of hidden units. Also, GRU outperforms LSTM on certain datasets, regardless of the choice units number; (ii) RNN variants yield a better accuracy than statistical or regression models; (iii) using ACF and PACF as dicovery tools in the feature selection process is unconclusive and unefficient when aiming for a general model; (iv) using deterministic features (such as day of the year, day of the month) has limited effects on improving the deep learning model’s performance. / Noggranna långsiktiga energiprognosprognoser för industriella enheter är av intresse för distributionsföretag eftersom det potentiellt kan bidra till att minska deras churn och erbjuda stöd i beslutsfattandet vid säkring. Detta avhandlingsarbete presenterar olika metoder för att prognostisera energiförbrukningen för industriella enheter under en lång tids förutsägelsehorisont på 1 år. I synnerhet inkluderar det experiment med två varianter av de återkommande neurala nätverken, nämligen GRU och LSTM. Deras prestanda jämförs med traditionella metoder, nämligen MLR och SARIMA. Vidare fokuserar undersökningen på att skräddarsy modellen för återkommande neurala nätverk för att förbättra prestanda. Experimenten fokuserar på effekterna av olika modellarkitekturer. För det andra fokuserar den på att testa effekten av tidsrelaterat funktionsval som en extra ingång till RNN -nätverk. Specifikt undersökte den hur traditionella metoder som Exploratory Data Analysis, Autocorrelation och Partial Autocorrelation Funtions Plots kan bidra till prestanda för RNN -modellen. Det aktuella arbetet visar genom en empirisk studie av tre industriella datamängder att GRU -arkitektur är en kraftfull metod för den långsiktiga prognosuppgiften som överträffar ac LSTM på vissa scenarier. Jämfört med MLR -modellen uppnådde RNN en minskning av RMSE mellan 5 % upp till 10 %. De viktigaste resultaten inkluderar: (i) GRU -arkitekturen överträffar LSTM på datauppsättningar för industriell energiförbrukning jämfört med ett lägre antal dolda enheter. GRU överträffar också LSTM på vissa datauppsättningar, oavsett antalet valenheter; (ii) RNN -varianter ger bättre noggrannhet än statistiska modeller eller regressionsmodeller; (iii) att använda ACF och PACF som verktyg för upptäckt i funktionsvalsprocessen är otydligt och ineffektivt när man siktar på en allmän modell; (iv) att använda deterministiska funktioner (t.ex. årets dag, månadsdagen) har begränsade effekter på att förbättra djupinlärningsmodellens prestanda.

Airborne discrete-return lidar is an active remote sensing technology capable of obtaining accurate, fine-resolution three-dimensional measurements over large areas. Discrete-return lidar data produce three-dimensional object characterizations in the form of point clouds defined by precise x, y and z coordinates. The data also provide intensity values for each point that help quantify the reflectance and surface properties of intersected objects. These data features have proven to be useful for the characterization of many important forest attributes, such as standing tree biomass, height, density, and canopy cover, with new applications for the data currently accelerating. This dissertation explores three new applications for airborne discrete-return lidar data. The first application uses lidar-derived metrics to predict understory vegetation cover, which has been a difficult metric to predict using traditional explanatory variables. A new airborne lidar-derived metric, understory lidar cover density, created by filtering understory lidar points using intensity values, increased the coefficient of variation (R²) from non-lidar understory vegetation cover estimation models from 0.2-0.45 to 0.7-0.8. The method presented in this chapter provides the ability to accurately quantify understory vegetation cover (± 22%) at fine spatial resolutions over entire landscapes within the interior ponderosa pine forest type. In the second application, a new method for quantifying and locating snags using airborne discrete-return lidar is presented. The importance of snags in forest ecosystems and the inherent difficulties associated with their quantification has been well documented. A new semi-automated method using both 2D and 3D local-area lidar point filters focused on individual point spatial location and intensity information is used to identify points associated with snags and eliminate points associated with live trees. The end result is a stem map of individual snags across the landscape with height estimates for each snag. The overall detection rate for snags DBH ≥ 38 cm was 70.6% (standard error: ± 2.7%), with low commission error rates. This information can be used to: analyze the spatial distribution of snags over entire landscapes, provide a better understanding of wildlife snag use dynamics, create accurate snag density estimates, and assess achievement and usefulness of snag stocking standard requirements. In the third application, live above-ground biomass prediction models are created using three separate sets of lidar-derived metrics. Models are then compared using both model selection statistics and cross-validation. The three sets of lidar-derived metrics used in the study were: 1) a 'traditional' set created using the entire plot point cloud, 2) a 'live-tree' set created using a plot point cloud where points associated with dead trees were removed, and 3) a 'vegetation-intensity' set created using a plot point cloud containing points meeting predetermined intensity value criteria. The models using live-tree lidar-derived metrics produced the best results, reducing prediction variability by 4.3% over the traditional set in plots containing filtered dead tree points. The methods developed and presented for all three applications displayed promise in prediction or identification of unique forest attributes, improving our ability to quantify and characterize understory vegetation cover, snags, and live above ground biomass. This information can be used to provide useful information for forest management decisions and improve our understanding of forest ecosystem dynamics. Intensity information was useful for filtering point clouds and identifying lidar points associated with unique forest attributes (e.g., understory components, live and dead trees). These intensity filtering methods provide an enhanced framework for analyzing airborne lidar data in forest ecosystem applications. / Graduation date: 2013

Au Grand-Duché de Luxembourg, le développement de modèles opérationnels pour la prévision des rendements se heurte actuellement au problème de la non prise en compte de la diminution de la surface verte utile et de sa relation avec des processus biotiques et abiotiques incriminés en situation de production. Pourtant, il apparaît que lélaboration dun modèle reliant la perte de surface verte à la baisse de rayonnement absorbé est la façon la plus adéquate daméliorer les prévisions de perte de rendement aux champs. De nombreuses manières destimer les rendements existent et se basent sur différentes approches et méthodes. Lobjectif de ce travail est de proposer un modèle destimation des rendements basé sur létude de la dégradation de la surface verte foliaire du blé et du rayonnement intercepté par cette surface tout au long de sa dégradation. Lapproche envisagée consiste, dans un premier temps, à utiliser les deux principaux modèles existants qui décrivent linterception du rayonnement par les surfaces foliaires vertes avec des données issues dexpérimentations aux champs, au Grand-Duché de Luxembourg en 2006 et 2007. Différentes méthodes dobtention des principales données dentrée de ces modèles ont été comparées et leurs avantages ont été mis en évidence. Ces données sont le LAI (Leaf Area Index) et le pourcentage de surface foliaire verte. Un LAI de référence, obtenu à laide dune méthode danalyse dimages de feuilles a été validé et comparé à une méthode dobtention du LAI basée sur la mesure du rayonnement intercepté par le couvert ainsi quà une méthode fournissant la couverture verte du sol à partir dimages aériennes de courte distance. Il a été montré que le LAI issu de la mesure du rayonnement intercepté et la couverture verte du sol sont obtenus plus rapidement et pour de plus grandes surfaces, mais quils ne sont pas suffisamment corrélés au LAI de référence pour être utilisés afin dobtenir le LAI réel. Le pourcentage de la surface foliaire verte de référence a également été obtenu à laide de la méthode danalyse dimages de feuilles. La comparaison de celui-ci aux estimations visuelles du pourcentage de surface verte foliaire a montré que cette méthode est plus rapide, mais engendre une surestimation du pourcentage de surface foliaire verte. Une relation linéaire significative entre la couverture verte du sol par prise dimages aériennes et le pourcentage de surface foliaire verte a été obtenue. Une amélioration de la prise dimages aériennes de courte distance pourrait mener à une substitution du pourcentage de surface foliaire verte par la couverture verte du sol sur de grandes surfaces à lavenir. Les deux principaux modèles décrivant linterception du rayonnement par les surfaces foliaires vertes ont été utilisés avec le LAI et le pourcentage de surface foliaire verte de référence. Une simplification de ces modèles par lutilisation de la dernière ou des deux dernières strates foliaires à la place des trois dernières pour le pourcentage de surface verte a montré que la simplification ne menait pas à une amélioration des résultats dans la plupart des cas. Dautre part, une estimation des biais introduits en utilisant les pourcentages de surface verte issus de lestimation visuelle à la place des estimations par lanalyse dimages montre que lestimation visuelle introduit un biais allant jusquà 20%. La comparaison des deux modèles testés a mené à la sélection du modèle aux sorties fournissant la meilleure relation avec les rendements. Cest une relation linéaire simple entre les paramètres de la courbe décrivant lévolution des sorties du modèle dit du « calcul de la matière sèche » au cours de la saison de culture et le rendement qui a été retenue. Dans un deuxième temps, le modèle sélectionné a été utilisé avec des données issues dexpérimentations menées de lannée 2000 à 2005, afin dobtenir une relation linéaire plus stable entre les rendements et les sorties de ce modèle. La relation obtenue montre des résultats significatifs et expliquant plus de 66% des rendements si une variété au comportement atypique est exclue. Un effet significatif de lannée, du précédent et de la variété sur cette relation a été mis en évidence. Dans un troisième temps, laspect prédictif du modèle destimation des rendements basé sur la relation linéaire simple retenue a été étudié sur deux années de données extérieures aux années utilisées pour la construction de celui-ci. Les données dentrée nécessaires au fonctionnement du modèle ont dû être obtenues de manière prédictive, afin de réaliser des estimations du rendement à venir à partir de la floraison. Le modèle Proculture, basé sur la simulation de lévolution des symptômes de la septoriose, a permis dobtenir des estimations en prévision des pourcentages de surface verte, et le LAI a été considéré comme constant par variété dune année à lautre. Le modèle destimation utilisé a permis dobtenir des prévisions de rendement ~20% supérieures aux rendements réels./In the Grand Duchy of Luxembourg, the development of operational models for predicting yields currently runs against the failure to take into account the green leaf area decline and its relationship with biotic and non biotic processes involved in a situation of production. Yet it appears that the development of a model linking the loss of green leaf area to lower radiation absorbed is the most adequate to improve prediction of yield loss in the fields. Many ways to estimate yields exist and are based on different approaches and methods. The objective of this work is to propose a model for estimating yields based on the study of the green leaf area decline of wheat and radiation intercepted by this area throughout the season. The approach is, first, to use the two main existing models that describe the interception of radiation by green leaf area with data from experiments in the field, in the Grand Duchy of Luxembourg in 2006 and 2007. Different methods for obtaining key data entry of these models were compared and their benefits have been identified. These data are LAI (Leaf Area Index) and the percentage of green leaf area. The reference method, obtained using image analysis of leaves has been validated and compared to a method for obtaining LAI based on the measurement of radiation intercepted by the canopy as well as a method based on the green cover soil obtained from short distance aerial images. It was shown that the LAI obtained from the measurement of radiation intercepted and the green land cover obtained from short distance aerial images are obtained faster and for larger surfaces, but they are not sufficiently correlated with the LAI from the reference method to be used in place of reference LAI. The percentage of green leaf area of reference has also been obtained using the image analysis of leaves. Comparing it to visual estimates of the percentage of green leaf area has shown that this method is faster and creates an overestimation of the percentage of green leaf area. A significant linear relationship between green land cover from short distance aerial images analysis and the percentage of green leaf area was obtained. An improved short distance aerial image could lead to the substitution of the percentage of green leaf area by the green land cover over large areas in the future. The two main models describing the interception of radiation by green leaf area were used with the LAI and the percentage of green leaf area of reference. A simplification of these models by using only the upper leaf or the two last leaves to emerge in place of the last three leaves to emerge for the percentage of green area has shown that simplification did not lead to improved results in most cases. On the other hand, an estimate of bias using the percentage of green leaf area from the visual estimate in place of estimates by image analysis shows that visual estimate introduce an approximate bias of 20%. A comparison of the two models tested led to the selection of the model outputs providing the best relationship with yields. It is a simple linear relationship between parameters of the curve describing the evolution of model outputs socalled calculation of dry matter during the growing season and yield that was chosen. In a second time, the selected model was used with data from experiments conducted from 2000 to 2005 to obtain a more stable linear relationship between yields and output of the model. The relationship obtained shows significant results and explains over 66% yields if datas from an atypical variety are excluded. A significant effect of years, precedent and variety on this relationship was highlighted. In a third time, the predictive aspect of the model to estimate yields based on the simple linear relationship has been studied on two years of external data used for years to build it. The input data needed to run the model had to be obtained on a predictive way to make estimates of future performance from flowering. The model Proculture, based on the simulation of the progression of septoriose disease, allowed obtaining estimates in anticipation of the percentage of green area, and LAI was considered constant variety from one year to another. The estimation model used resulted in expected future performance ~ 20% higher than actual yields.

Energimyndigheten har drivit ett projekt kallat ”Förbättrad energistatistik i bebyggelsen” för att få mer kunskap om energianvändningen i byggnader.  Denna rapport fokuserar på ”Mätning av hushållsel på apparatnivå” som var ett delprojekt. Diverse regressionsmodeller används i denna rapport för att undersöka sambandet mellan elanvändningen och de olika förklarande variablerna, som exempelvis hushållens bakgrundsvariabler, hushållstyp och geografiska läge, elförbrukningen av olika elapparater samt antalet elapparater. Datamaterialet innefattar 389 hushåll där de flesta är spridda runt om i Mälardalen. Ett fåtal mätningar gjordes på hushåll i Kiruna och Malmö. Slutsatsen vi kan dra från denna uppsats är att hushållens bakgrund, hustyp, geografiska läge och antal elapparater samt dessa apparaters typ har relevans för elförbrukningen i ett hushåll. / The Swedish Energy Agency conducted a project which is called “Improved energy statistics for settlements”. This report focuses on one field of the project: “households’ electricity use on device level”. Various regression models are used in the analysis to analyze the relationship between electricity usage and different explanatory variables, for instance: background variables for the household, type of household, geographical setting, usage of different electrical devices and quantity of electrical devices used.  The data material consists of 389 households which are spread around the region of Märlardalen except for a few households from the communities of Kiruna and Malmö. The conclusion we can draw from this thesis shows that the background variables for a household, its type, its geographical setting and the amount and type of devices it contains all have a contribution to the electricity usage in the household. / Förbättrad energistatistik i bebyggelsen

There has been increasing interest in evaluating the performance of existing reinforced concrete (RC) bridges just after natural disasters or man-made events especially when the defects are invisible, or in quantifying the improvement after rehabilitations. In order to obtain an accurate assessment of the reliability of a RC bridge, it is critical to incorporate information about its current structural properties, which reflects the possible aging and deterioration. This dissertation proposes to develop an adaptive reliability analysis of RC bridges incorporating the damage detection information obtained from nondestructive testing (NDT). In this study, seismic fragility is used to describe the reliability of a structure withstanding future seismic demand. It is defined as the conditional probability that a seismic demand quantity attains or exceeds a specified capacity level for given values of earthquake intensity. The dissertation first develops a probabilistic capacity model for RC columns and the capacity model can be used when the flexural stiffness decays nonuniformly over a column height. Then, a general methodology to construct probabilistic seismic demand models for RC highway bridges with one single-column bent is presented. Next, a combination of global and local NDT methods is proposed to identify in-place structural properties. The global NDT uses the dynamic responses of a structure to assess its global/equivalent structural properties and detect potential damage locations. The local NDT uses local measurements to identify the local characteristics of the structure. Measurement and modeling errors are considered in the application of the NDT methods and the analysis of the NDT data. Then, the information obtained from NDT is used in the probabilistic capacity and demand models to estimate the seismic fragility of the bridge. As an illustration, the proposed probabilistic framework is applied to a reinforced concrete bridge with a one-column bent. The result of the illustration shows that the proposed framework can successfully provide the up-to-date structural properties and accurate fragility estimates.

Nonnegative matrix factorization (NMF) is a useful dimension reduction method that has been investigated and applied in various areas. NMF is considered for high-dimensional data in which each element has a nonnegative value, and it provides a low-rank approximation formed by factors whose elements are also nonnegative. The nonnegativity constraints imposed on the low-rank factors not only enable natural interpretation but also reveal the hidden structure of data. Extending the benefits of NMF to multidimensional arrays, nonnegative tensor factorization (NTF) has been shown to be successful in analyzing complicated data sets. Despite the success, NMF and NTF have been actively developed only in the recent decade, and algorithmic strategies for computing NMF and NTF have not been fully studied. In this thesis, computational challenges regarding NMF, NTF, and related least squares problems are addressed. First, efficient algorithms of NMF and NTF are investigated based on a connection from the NMF and the NTF problems to the nonnegativity-constrained least squares (NLS) problems. A key strategy is to observe typical structure of the NLS problems arising in the NMF and the NTF computation and design a fast algorithm utilizing the structure. We propose an accelerated block principal pivoting method to solve the NLS problems, thereby significantly speeding up the NMF and NTF computation. Implementation results with synthetic and real-world data sets validate the efficiency of the proposed method. In addition, a theoretical result on the classical active-set method for rank-deficient NLS problems is presented. Although the block principal pivoting method appears generally more efficient than the active-set method for the NLS problems, it is not applicable for rank-deficient cases. We show that the active-set method with a proper starting vector can actually solve the rank-deficient NLS problems without ever running into rank-deficient least squares problems during iterations. Going beyond the NLS problems, it is presented that a block principal pivoting strategy can also be applied to the l1-regularized linear regression. The l1-regularized linear regression, also known as the Lasso, has been very popular due to its ability to promote sparse solutions. Solving this problem is difficult because the l1-regularization term is not differentiable. A block principal pivoting method and its variant, which overcome a limitation of previous active-set methods, are proposed for this problem with successful experimental results. Finally, a group-sparsity regularization method for NMF is presented. A recent challenge in data analysis for science and engineering is that data are often represented in a structured way. In particular, many data mining tasks have to deal with group-structured prior information, where features or data items are organized into groups. Motivated by an observation that features or data items that belong to a group are expected to share the same sparsity pattern in their latent factor representations, We propose mixed-norm regularization to promote group-level sparsity. Efficient convex optimization methods for dealing with the regularization terms are presented along with computational comparisons between them. Application examples of the proposed method in factor recovery, semi-supervised clustering, and multilingual text analysis are presented.

Processor architectures are becoming increasingly complex and hence architects have to evaluate a large design space consisting of several parameters, each with a number of potential settings. In order to assist in guiding design decisions we develop simple and accurate models of the superscalar processor design space using a detailed and validated superscalar processor simulator. Firstly, we obtain precise estimates of all significant micro-architectural parameters and their interactions by building linear regression models using simulation based experiments. We obtain good approximate models at low simulation costs using an iterative process in which Akaike’s Information Criteria is used to extract a good linear model from a small set of simulations, and limited further simulation is guided by the model using D-optimal experimental designs. The iterative process is repeated until desired error bounds are achieved. We use this procedure for model construction and show that it provides a cost effective scheme to experiment with all relevant parameters. We also obtain accurate predictors of the processors performance response across the entire design-space, by constructing radial basis function networks from sampled simulation experiments. We construct these models, by simulating at limited design points selected by latin hypercube sampling, and then deriving the radial neural networks from the results. We show that these predictors provide accurate approximations to the simulator’s performance response, and hence provide a cheap alternative to simulation while searching for optimal processor design points.

Studentų pažangą geriausiai nusako jų studijų dalykų bei bakalauro darbo vertinimai. Šiaulių universitete Matematikos ir informatikos fakultete Matematikos bei Matematikos ir informatikos bakalauro studijų programose pagrindiniai studijų dalykai yra: Matematinė analizė, Algebra, Fizika, Geometrija, Kompiuterinės matematikos sistemos ir Programavimo pagrindai. Bakalauro darbe analizuojame Šiaulių universiteto Matematikos ir informatikos fakulteto Matematikos bei Matematikos ir informatikos bakalauro studijų programų studentų įstojusių 2004–2008 pasiekimus. Šio darbo tikslas yra nustatyti, kuris iš pagrindinių studijų dalykų vertinimų labiausiai susijęs su baigiamojo bakalauro darbo vertinimu. Šiam ryšiui nustatyti naudojame koreliacinę analizę bei tiesinę regresiją. / Students’ advancement is described by their evaluation of study subjects and Bachelor's work. Basic subjects in Bachelor's degree program at Siauliai University, Faculty of Mathematics and Informatics are: Mathematical analysis, Algebra, Geometry, Physics, Computer mathematics system and Programming fundamentals. In Bachelor's work, we investigate a relation between 6 study subjects and Bachelor's work evaluations of students who entranced in 2004–2008. To describe this relation we use correlation analysis and linear regression.