Conhecer a resposta da cultura do algodoeiro (Gossypium hirsutum L var. latifolium Hutch) à variabilidade espacial de atributos físicos e químicos do solo e seu comportamento espectral mediante doses variadas de fertilizantes pode ajudar a estabelecer estratégias de manejo do solo e da cultura, com melhor uso de corretivos e fertilizantes. Dentro deste contexto, este trabalho foi desenvolvido para: a) estudar a variabilidade espacial da fertilidade do solo, estado nutricional, população de plantas, número de estruturas reprodutivas e produtividade do algodoeiro em uma lavoura comercial; b) verificar as relações de causa e efeitos entre as variáveis estudadas; c) avaliar o efeito da variação de doses de nitrogênio na detecção do teor de nitrogênio foliar do algodoeiro utilizando propriedades de reflectância através de um sensor óptico ativo. Para isso, foram conduzidos dois experimentos diferenciados. O primeiro foi realizado numa lavoura comercial, localizada em Chapadão do Céu, GO. Os dados foram coletados numa área de 90 hectares, dividida numa malha com espaçamento regular de 100 m entre pontos, formando um retângulo com 10 colunas e 11 linhas. Para detectar a variação na pequena escala, foram feitas amostragens mais intensas com espaçamento de 33 m, formando 5 "ilhas", constituindo um total de 170 amostras. Os dados foram submetidos às análises estatísticas descritivas, geoestatística, interpolação através de krigagem ordinária e regressão pelo método "stepwise". A variabilidade expressa pelo coeficiente de variação foi baixa a moderada para todas as variáveis analisadas, com exceção de maçãs retidas e P no solo. Os resultados da análise geoestatística indicaram que os fatores de produção da cultura do algodoeiro apresentam, em sua maioria, dependência espacial e esta deve ser considerada quando da definição de esquemas de amostragem e práticas de manejo do solo e da cultura. A produtividade do algodoeiro foi significativamente afetada por N, P, Ca e Mg no tecido foliar e fração areia e argila do solo. O segundo experimento foi realizado na área experimental da USP/ESALQ, em Piracicaba, SP. O delineamento experimental foi o aleatorizado em blocos, com cinco tratamentos (0, 50, 100, 150 e 200 kg ha-1 de N) e quatro repetições. Os dados do índice de vegetação da diferença normalizada (NDVI) e da razão simples vermelho/infravermelho próximo (V/IVP) foram coletados no período de 23 a 92 dias após a emergência (DAE) totalizando nove datas de coleta. Foram coletados também dados referentes ao índice de clorofila, índice de área foliar (IAF) e altura de plantas. Os valores do NDVI apresentaram uma tendência crescente ao longo do período das leituras, enquanto a relação V/IVP teve comportamento inverso, e a partir de 45 DAE, foram afetados significativamente pelas doses de N aplicadas. Relações significativas foram observadas entre os valores do NDVI e índice de clorofila e IAF. Aos 92 DAE o NDVI apresentou relação linear positiva e altamente significativa com o teor foliar de N. Os resultados obtidos indicam que as técnicas de geoestatística e sensoriamento remoto podem auxiliar na tomada de decisões sobre as práticas de calagem e adubação em taxa variada na cultura do algodoeiro. / Know the cotton answer (Gossypium hirsutum L. var. latifolium Hutch) to the spatial variability of soil physical and chemical attributes and its spectral behavior by means of varied rates of nutrients can help to establish soil and crop management strategies, with better use of lime and fertilizers. This work was developed to: a) study the spatial variability of soil fertility, nutritional state, plants population, reproductive structures number and cotton yield in a commercial field; b) to study the relationships among those variables; c) to evaluate the effect of the variation of nitrogen levels on the detection of its content on cotton leaves using reflectance properties by an active optical sensor. Two differentiated experiments were conducted; the first one was located in a commercial field in Chapadão do Céu, GO. Data were collected in a area of 90 hectares, divided in a regular grid of 100 m with 10 columns and 11 lines. In order to detect the variation in the small scale, more intense samplings with spacing of 33 m were conducted forming 5 islands, constituting a total of 170 samples. The data were submitted to the descriptive statistical analysis, geostatistic, interpolation through ordinary kriging, and stepwise regression analysis. The variability expressed by variation coefficient were low to moderated for all analyzed variables, except for healthy bolls and soil P. Geostatistics analysis indicated that the crop cotton factors of production present, in its majority, spatial dependence and this should be considered when of the definition of sampling schemes and soil and crop management practices. Cotton yield was affected significantly by N, P, Ca and Mg in the leaves, as well as sand and clay fraction of the soil. The second experiment was accomplished in an experimental area at USP/ESALQ, in Piracicaba, SP. The experiment was conducted as a randomized block design and treatments consisted on five levels of nitrogen fertilization (0, 50, 100, 150 and 200 kg of N ha-1) with four replications. The readings of Normalized Difference Vegetation Index (NDVI and the ratio Red/Near infrared (RED/NIR) were collected between 23 to 92 days after the emergency (DAE), with the total of 9 visits. Data referred to the chlorophyll index, leaf area index (LAI) and plants height were also collected. The values of NDVI presented a growing tendency along the period, while RED/NIR had an inverse behavior, being significantly affected by the rates of applied N, starting from 45 DAE. At 92 DAE, NDVI presented a positive linear relationship with leaves N content. Significant relationships were observed between the values of NDVI and chlorophyll index and LAI. The results obtained indicate that geostatistic techniques and remote sensing can help on management decisions for variable rate application of lime and fertilizers in cotton.

In the last two decades, energy is becoming one of the main issues in the manufacturing industry as it contributes substantially to production cost, CO2 emissions, and other destructive environmental impact. Due to rising energy costs, environmental concerns and stringent regulations, manufacturing is increasingly driven towards sustainable manufacturing which needs to address the associated environmental, social and economic aspects simultaneously. One common approach is to achieve sustainability and to implement energy-resource efficient production management systems that enable optimisation of energy consumption and resource utilisation in the production system. However, by reducing energy consumption, the product quality and production cost may be compromised. To remain competitive in the dynamic environment, the energy-efficient management system should not only concern energy consumption but also maintain product quality and production efficiency. This thesis presents a development of the Energy-smart Production Management (e-ProMan) system which provides a systematic, virtual simulation that integrates manufacturing data relating to thermal effect and correlation analysis between energy flow, work flow and data flow for the heating, ventilation and air conditioning (HVAC) system and production process. First, the e-ProMan system comprises of the multidimensional analysis between energy flow, work flow and data flow. The results showed that the product quality is significantly affected by ambient temperature in CNC precision machining. Product quality appears to be improved at lower temperatures. This research highlights the significance of ambient temperature in sustainable precision machining. Second, the simulation experiment was modelled at the production process due to it being the main source of energy consumption in manufacturing. An up-hill workload scenario was found to be the most energy and cost-efficient production processes. In other words, energy consumption, CO2 emission and total manufacturing cost could be reduced when workload capacity and operating machine increase incrementally. Moreover, the e-ProMan system was modelled and simulated using the weather forecast and real-time ambient temperature to reduce energy consumption of the HVAC system. The e-ProMan system results in less energy consumption compared to the fuzzy control system. To conclude, the e-ProMan demonstrates energy efficiency at all relevant levels in the manufacturing: machine, process and plant. For the future research, the e-ProMan system needs to be applied and validated in actual manufacturing environments.

This thesis provides novel methodology for statistical analysis of paired high-dimensional genomic data, with the aimto identify gene interactions specific to each group of samples as well as the gene connections that change between the two classes of observations. An example of such groups can be patients under two medical conditions, in which the estimation of gene interaction networks is relevant to biologists as part of discerning gene regulatory mechanisms that control a disease process like, for instance, cancer. We construct these interaction networks fromdata by considering the non-zero structure of correlationmatrices, which measure linear dependence between random variables, and their inversematrices, which are commonly known as precision matrices and determine linear conditional dependence instead. In this regard, we study three statistical problems related to the testing, single estimation and joint estimation of (conditional) dependence structures. Firstly, we develop hypothesis testingmethods to assess the equality of two correlation matrices, and also two correlation sub-matrices, corresponding to two classes of samples, and hence the equality of the underlying gene interaction networks. We consider statistics based on the average of squares, maximum and sum of exceedances of sample correlations, which are suitable for both independent and paired observations. We derive the limiting distributions for the test statistics where possible and, for practical needs, we present a permuted samples based approach to find their corresponding non-parametric distributions. Cases where such hypothesis testing presents enough evidence against the null hypothesis of equality of two correlation matrices give rise to the problem of estimating two correlation (or precision) matrices. However, before that we address the statistical problem of estimating conditional dependence between random variables in a single class of samples when data are high-dimensional, which is the second topic of the thesis. We study the graphical lasso method which employs an L1 penalized likelihood expression to estimate the precision matrix and its underlying non-zero graph structure. The lasso penalization termis given by the L1 normof the precisionmatrix elements scaled by a regularization parameter, which determines the trade-off between sparsity of the graph and fit to the data, and its selection is our main focus of investigation. We propose several procedures to select the regularization parameter in the graphical lasso optimization problem that rely on network characteristics such as clustering or connectivity of the graph. Thirdly, we address the more general problem of estimating two precision matrices that are expected to be similar, when datasets are dependent, focusing on the particular case of paired observations. We propose a new method to estimate these precision matrices simultaneously, a weighted fused graphical lasso estimator. The analogous joint estimation method concerning two regression coefficient matrices, which we call weighted fused regression lasso, is also developed in this thesis under the same paired and high-dimensional setting. The two joint estimators maximize penalized marginal log likelihood functions, which encourage both sparsity and similarity in the estimated matrices, and that are solved using an alternating direction method of multipliers (ADMM) algorithm. Sparsity and similarity of thematrices are determined by two tuning parameters and we propose to choose them by controlling the corresponding average error rates related to the expected number of false positive edges in the estimated conditional dependence networks. These testing and estimation methods are implemented within the R package ldstatsHD, and are applied to a comprehensive range of simulated data sets as well as to high-dimensional real case studies of genomic data. We employ testing approaches with the purpose of discovering pathway lists of genes that present significantly different correlation matrices on healthy and unhealthy (e.g., tumor) samples. Besides, we use hypothesis testing problems on correlation sub-matrices to reduce the number of genes for estimation. The proposed joint estimation methods are then considered to find gene interactions that are common between medical conditions as well as interactions that vary in the presence of unhealthy tissues.

A qualidade e a produtividade agrícolas estão altamente relacionadas com a disponibilidade de água, já que sua falta pode causar prejuízos significativos em períodos críticos. Por esse motivo, a irrigação torna-se uma maneira importante de manter a disponibilidade de água, com o objetivo de obtenção de altos índices de qualidade e produtividade agrícolas. Os mapas temáticos desempenham papel de fundamental importância no processo de tomada de decisão em agricultura de precisão. Para a geração de mapas desse tipo, estão disponíveis bibliotecas e interfaces de código livre, como a TerraLib, a SOMCode e a aRT, que possuem algoritmos capazes de realizar a interpolação de dados com boa precisão, além de agrupar dados em regiões com características semelhantes. No presente trabalho foram utilizados dados obtidos por uma grade de sensores de uma unidade piloto do sistema Irrigap, desenvolvida pela Embrapa Instrumentação Agropecuária, dados esses referentes a medidas horárias de umidade e temperatura do solo e temperatura do ar. Diferentes métodos de interpolação foram utilizados, com o intuito de verificar qual a melhor maneira de se gerar mapas que possibilitem a visualização de áreas com o mesmo comportamento com relação às variáveis de solo e climatológicas, as chamadas zonas de manejo agrícolas. Os resultados obtidos mostraram que é possível a identificação dessas zonas, desde que sejam utilizados os algoritmos corretos para cada situação. Além disso, foi observado que softwares para geração de mapas temáticos utilizando dados reais podem se tornar importantes para a tomada de decisão em agricultura de precisão e proporcionar benefícios como a economia de água e energia durante o período de irrigação. / The agricultural quality and yield are highly related with the water readiness, because its privation can cause significant damages in critical periods. Due to that reason, irrigation becomes an important way of readiness water maintenance, with the goal of obtaining high level of agricultural quality and yield. The thematic maps have a very important role in the precision agriculture making decision. For that kind of maps generation, there are free code libraries and interfaces, like TerraLib, SOMCode and aRT, whit algorithms to accomplish data interpolation with reliable precision, and also to group data from areas with similar behavior. In the present work, data were obtained from sensor grid of a pilot unit of the Irrigap system, developed by Embrapa Instrumentação Agropecuária, referring to hourly measures of soil moisture and temperature and air temperature. Different interpolation methods were used, with the intention of verify the best way to generate maps to make possible the visualization of areas with the same behavior related to the soil and the climatological variables, know as management zones. Obtained results show that the identification of these zones is possible, since the correct algorithms are used for each situation. Besides that, was observed that the software developed to generate thematic maps using real data can be important for making decision in precision agriculture and to provide benefits like saving water and energy during the irrigation period.

The purpose of this thesis is to analyse studies that evaluate Web search engines. This is done in four categories; the researchers’ purpose, the evaluation measurements, the relevance, and the time aspect. Our method is based on a text analysis in which we use the direction of analysis of the content of sixteen evaluation experiments. Our results indicate fundamental differences in the way the researchers are tackling the problem of evaluation of Web search engines. We think that, despite the differences that we have been able to identify, it is necessary to perform evaluation experiments, so that methods can be developed that can guarantee the quality of the Web search engines. To provide people with the kind of information they need is the main task for Web search engines. In an increasing flow of information that task will be even more important. Evaluation of Web search engines can be a part of improving the efficiency of the Web search engines and in that way strengthen their roll as important information resources. / Uppsatsnivå: D

This thesis is focusing on how global and local search engines retrieve information from the local domain. The three search engines tested are the global search engines Yahoo, Ask.com and the local search engine Web Wombat. The questions we examined were: which search engine has the best retrieval effectiveness? Could there be other reasons than retrieval effectiveness, to choosing a local search engine?For our test we constructed 20 questions related to Australia.We chose to divide the questions into topics like nature, sports and culture. For all questions we evaluated the relevance of the first 20 hits. We used the following measures in our test: Jaccard´s index, precision and average precision. We also looked at factors such as duplicates and error pages, because we consider this to be an important aspect to consider, when looking at the relevance of the first 20 hits. The results of our study showed that Yahoo had the best performance for precision. Web Wombat had faulty precision but results from Jaccard´s index revealed that Web Wombat had many unique documents. Web Wombat had the best average precision on one of our questions. In spite of Web Wombats faulty precision, we think that Web Wombat serve a purpose as an alternative to global search engines. / Uppsatsnivå: D

The purpose of this Master’s thesis is to evaluate three general query-based web search engines. The search engines chosen are Google, Yahoo! and Live Search. The topics, which where transformed into twenty queries, are based on real users’ information needs, taken from an ask-a-question service. For every search the first 10 hits in each of the three search engines were evaluated. The measures that have been used are precision and overlap. The queries were evaluated on a binary relevance scale with 0 or 1. Precision has been measured as average precision over 1-10 DCV levels for all 20 queries in the evaluation. It has also been measured as average precision in a query-by-query calculation over 10 DCV levels for each query and search engine. The number of duplicate links, dead links, mirror pages, relevant hits and irrelevant hits has also been recorded in the study for each search engine, for a comparison. Overlap was measured at two different DCVs, 1 and 5. Eventually the results show that Google had the best results regarding precision. It also had the highest number of relevant hits. Google and Yahoo! showed the highest overlap at DCV 1. At DCV 5 the greatest overlap was between Yahoo! and Live Search. / Uppsatsnivå: D

The purpose of this thesis is to evaluate and compare the retrieval effectiveness of two search engines on the Internet: Svensk Medicin and Google Scholar. Svensk Medicin is a medicine-specific search engine and Google Scholar specialises in indexing scholarly material. A set of 20 questions in four categories were used to conduct the searches with the search engines. Genuine medical information needs were selected from a question collection provided by the television program Fråga doktorn. The relevance of the first 10 retrieved documents was judged by using a binary scale and the measure used was precision. An analysis on average precision for all the search questions and an average measure was calculated (macro precision). Results for average precision over all the search questions and a mean average precision is also presented (micro precision and map). These measures shows which search engine performs best for single queries and how the relevant documents spread among the displayed results. The occurrence of duplicates and error pages was noted, because they also show how the search engine performs. Svensk Medicin is performing better than Google Scholar on the single queries and Svensk Medicin also retrieves the highest number of relevant documents. Methodology is discussed close since the experimental design affects the result. / Uppsatsnivå: D

The aim of this thesis was to evaluate the music downloading service Musikwebb regarding its indexing and retrieval effectiveness. This was done by performing various kinds of search in the system. The outcome of these searches were then analysed according to the criteria specificity, precision, recall, exclusivity and authority control. The study showed that Musikwebb had several flaws regarding its retrieval effectiveness. The most prominent cases were the criteria exclusivity and specificity. Several of Musikwebb’s classes could be regarded as almost similar and the average number of songs in each class was over 50 000. As this study shows, having over 50 000 unique entries in a class results in problems regarding the effectiveness of the browsing technique. The developers of Musikwebb are recommended by the author to acquire their licensed material from All Music Guide, including the implementation of the All Music Guide classification system.

Orientador: Célia Regina Lopes Zimback / Resumo: Com o crescente interesse de produtores rurais e grandes empresas no ramo da agricultura de precisão, muito tem sido feito e desenvolvido para adequar essa tecnologia, na sua maioria estrangeira, para as condições tropicais. Aplicadores de insumos à taxa variável são cada vez mais comuns, otimizando as operações e reduzindo seu impacto ambiental. Mas para que isso funcione de maneira dinâmica como ocorre no solo, é necessário conhecer a dependência espacial entre os parâmetros analisados, aplicando a geoestatística uni e multivariada no lugar da estatística tradicional. Assim, mapas mais condizentes com a realidade podem ser elaborados por krigagem, com auxílio de um Sistema de Informações Geográficas (SIG). O objetivo desse trabalho foi investigar o comportamento de variáveis ligadas à química do solo por meio de técnicas geoestatísticas, a fim de entendê-lo e obter mapas de fertilidade da área, auxiliando na elaboração de um mapa final de aplicação de fertilizantes em formato digital. A área estudada foi de 71,79 ha em uma propriedade rural localizada em Araguari-MG, utilizando-se como metodologia amostragens com grade conhecida e amostras georreferenciadas. Pôde-se concluir que a amostragem em malha (grid) regular de 60 m possibilitou a caracterização geoestatísticas das variáveis estudadas da fertilidade do solo, exceto para P e K; a análise de corregionalização foi uma ferramenta eficaz no estudo do comportamento das variáveis envolvidas em diferentes processos no solo; a krigagem mostrou-se um interpolador eficaz na confecção de mapas de fertilidade quando detectada a dependência espacial entre as amostras; quando não detectada a dependência espacial, o Inverso do Quadrado da Distância foi uma boa alternativa de interpolação para a construção de mapas de fertilidade;...(Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Following the great interest of farmers and big companies in precision farming, lot of things have been done to adapt this technology, in majority foreigner, for tropical conditions. Agrotoxics variable rate applicators are each time more commons, optimizing operations and reducing environmental impact. But to make that work dynamically as the soil as, is necessary to know spatial dependence among analyed parameters, using univariate and multivariate geostatistics instead of traditional statistics. In this way, maps can be elaborate with Geografical Informations Systems (GIS). The area studied covers 71,79 ha of a rural property at Araguari-MG, in which georeferrenced samples were taken allow a known grid. It could be concluded that 60m grid soil sampling allowed geostatistic characterization of the studied soil fertility variables, excepted P and K; corregionalization analysis was a efficient tool on studied variables behavior involved in different soil process; kriging has proved to be a efficient interpolator for fertility maps when there was spatial correlation; otherwise inverse distance weight was used as an good alternative for fertility maps; finally Geographic Information System (GIS) optimize data management process, helping maps preparation for the prescription of fertilizers and liming. / Mestre