Submitted by Neusa Fagundes (neusa.fagundes@unioeste.br) on 2018-06-15T20:19:22Z No. of bitstreams: 2 Fabiane_Fontana2018.pdf: 2677532 bytes, checksum: 3036328537227cc96b8ea368e893f2fc (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2018-06-15T20:19:22Z (GMT). No. of bitstreams: 2 Fabiane_Fontana2018.pdf: 2677532 bytes, checksum: 3036328537227cc96b8ea368e893f2fc (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-07-19 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Precision Agriculture (AP) uses technologies aimed at increasing productivity and reducing environmental impact through localized application of agricultural inputs. In order to make AP economically feasible, it is essential to improve current methodologies, as well as to propose new ones, such as the design of management areas (MZs) from productivity data, topographic, and soil attributes, among others, to determine which are heterogeneous subareas among themselves in the same area. In this context, the main objective of this research was to evaluate three distance metrics (Diagonal, Euclidian, and Mahalanobis) through FUZME and SDUM software (for the definition of management units) using the fuzzy c-means algorithm, and, at a further moment, to evaluate the cultures of soybeans and corn, as well as the association between them. On the first scientific paper, using data corresponding to four distinct areas, the three metrics with original and normalized data associated with soybean yield were evaluated. For area A, the Diagonal and Mahalanobis distances exempted the need for normalization of the variables, presenting areas that were identical for both versions. After the normalization of the data, the Euclidian distance presented a better delineation in its MZs for area A. For areas B, C, and D it was not possible to reach conclusions regarding the best performance, since only one variable was used for the process of MZs, and that has directly influenced the results. On the second scientific paper, data corresponding to three distinct areas were applied to analyze the use of soybean and corn yields, as well as the association between them, in the selection of variables to define MZs. Based on the variables available for each of the areas, the selection was carried out using the spatial correlation method, considering, for each one of the areas, the three target yields (soybean, corn, and soybean+corn). The type of productivity used demonstrated two different outcomes: first in the variable selection process, where its alternation resulted in different selections for the same area, and second, in the evaluation of the defined MZs, where even when the same variables were selected in the definition of the MZs, the performances of the MZs were different. After the validation methods applied, it was verified that the best target yield was soybean+corn, reasserting the idea of being better to use these two cultures, together, when defining the MZs of an area with rotating crops of soybean and corn. / A Agricultura de Precisão (AP) utiliza tecnologias objetivando o aumento da produtividade e redução do impacto ambiental por meio de aplicação localizada de insumos agrícolas. Para viabilizar economicamente a AP, é essencial aprimorar as metodologias atuais, bem como propor novas, como, por exemplo, o delineamento de zonas de manejo (ZMs) a partir de dados de produtividade, atributos topográficos e do solo, entre outros, utilizados a fim de determinar subáreas heterogêneas entre si em uma mesma área. Neste contexto, este trabalho teve como principal objetivo avaliar três métricas de distâncias (Diagonal, Euclidiana e Mahalanobis) junto aos Softwares FUZME e SDUM (Software para a definição de unidades de manejo), que utilizam o algoritmo fuzzy c-means, e, em um segundo momento, avaliar também as culturas de soja e milho, assim como a associação entre elas. No primeiro artigo, utilizando dados correspondentes a quatro áreas distintas, avaliaram-se as três métricas com dados originais e normalizados associados à produtividade de soja. Para a área A, as distâncias Diagonal e Mahalanobis dispensaram a necessidade de normalização das variáveis, apresentando áreas idênticas para as duas versões. Após a normalização dos dados, a distância Euclidiana apresentou um melhor delineamento em suas ZMs para a área A. Para as áreas B, C e D não foi possível obter conclusões quanto ao melhor desempenho, visto que o fato de ser utilizado apenas uma variável para o processo de definição de ZMs influenciou diretamente nos resultados obtidos. No segundo artigo, dados correspondentes a três áreas distintas foram utilizados para analisar o uso de produtividades de soja e milho, assim como a associação entre elas, na seleção de variáveis para definição de ZMs. A partir das variáveis disponíveis para cada uma das áreas foi realizada a seleção destas através do método da correlação espacial, levando em consideração, para cada uma das áreas, as três produtividades-alvo (soja, milho e soja+milho). O tipo de produtividade utilizada repercutiu de duas formas diferentes: primeiro no processo de seleção de variáveis, onde a sua alternância resultou em seleções diferenciadas para uma mesma área; e em um segundo momento, na avaliação das ZMs definidas, onde mesmo quando as mesmas variáveis foram selecionadas na definição das ZMs, os desempenhos das ZMs foram diferentes. Após os métodos de validação aplicados, verificou-se que a melhor produtividade-alvo foi soja+milho, reforçando a ideia de ser útil a utilização destas duas culturas, em conjunto, na definição das ZMs de uma área com alternância de produção de soja e milho.

Perioperative hypotension (PH), commonly a side effect of anesthesia,is one of the main mortality causes during the 30 posterior days of asurgical procedure. Novel research lines propose combining machinelearning algorithms with the Arterial Blood Pressure (ABP) waveform tonotify healthcare professionals about the onset of a hypotensive event withtime advance and prevent its occurrence. Nevertheless, ABP waveformsare heterogeneous among patients, consequently, a general model maypresent different predictive capabilities per individual. This project aimsat improving the performance of an artificial neural network (ANN) topredict hypotension events with time advance by applying personalizedmachine learning techniques, like data grouping and domain adaptation. Wehypothesize its implementation will allow us to cluster patients with similardemographic and ABP discriminative characteristics and tailor the modelto each specific group, resulting in a worst overall but better individualperformance. Results present a slight but not clinical significant improvementwhen comparing AUROC values between the group-specific and the generalmodel. This suggests even though personalization could be a good approach todealing with patient heterogeneity, the clustering algorithm presented in thisthesis is not sufficient to make the ANN clinically feasible. / Perioperativ hypotoni (PH), vanligtvis en sidoeffekt av anestesi, är en av dehuvudsakliga dödsorsakerna under de första 30 dagarna efter ett kirurgiskt ingrepp. Nya forskningslinjer föreslår att kombinera maskininlärningsalgo-ritmer med vågformen av det arteriella blodtrycket (ABP) för att förvarna sjukvårdspersonalen om uppkomsten av en hypotensiv episod, and därmedförhindra förekomsten. ABP-vågformen är dock heterogen bland patienter,så en allmän modell kan ha olik prediktiv förmåga för olika individer.I det här projektet används personaliserade maskininlärningstekniker, somdatagruppering och domänanpassning, för att försöka förbättra ett artificielltneuralt nätverk (ANN) som förutspår hytotensiva episoder. Vår hypotes är attimplementeringen kommer låta oss klustra patienter med liknande demografioch ABP-karakteristik för att skräddarsy modellen till varje specifik grupp,vilket leder till en sämre övergripande men bättre individuell prestanda. Resultaten visar små men inte kliniskt signifikanta förbättringar när AUROC-värden jämförs mellan den gruppspecifika och den allmänna modellen. Detta tyder på att även fast personalisering kan vara en bra tillnärmning till patientersheterogenitet, är inte klusteralgoritmen som presenteras här tillräcklig förklinisk användning av ANN.

Cette thèse aborde le processus ponctuel déterminantal, un modèle probabiliste qui capture la répulsion entre les points d’un certain espace. Celle-ci est déterminée par une matrice de similarité, la matrice noyau du processus, qui spécifie quels points sont les plus similaires et donc moins susceptibles de figurer dans un même sous-ensemble. Contrairement à la sélection aléatoire uniforme, ce processus ponctuel privilégie les sous-ensembles qui contiennent des points diversifiés et hétérogènes. La notion de diversité acquiert une importante grandissante au sein de sciences comme la médecine, la sociologie, les sciences forensiques et les sciences comportementales. Le processus ponctuel déterminantal offre donc une alternative aux traditionnelles méthodes d’échantillonnage en tenant compte de la diversité des éléments choisis. Actuellement, il est déjà très utilisé en apprentissage automatique comme modèle de sélection de sous-ensembles. Son application en statistique est illustrée par trois articles. Le premier article aborde le partitionnement de données effectué par un algorithme répété un grand nombre de fois sur les mêmes données, le partitionnement par consensus. On montre qu’en utilisant le processus ponctuel déterminantal pour sélectionner les points initiaux de l’algorithme, la partition de données finale a une qualité supérieure à celle que l’on obtient en sélectionnant les points de façon uniforme. Le deuxième article étend la méthodologie du premier article aux données ayant un grand nombre d’observations. Ce cas impose un effort computationnel additionnel, étant donné que la sélection de points par le processus ponctuel déterminantal passe par la décomposition spectrale de la matrice de similarité qui, dans ce cas-ci, est de grande taille. On présente deux approches différentes pour résoudre ce problème. On montre que les résultats obtenus par ces deux approches sont meilleurs que ceux obtenus avec un partitionnement de données basé sur une sélection uniforme de points. Le troisième article présente le problème de sélection de variables en régression linéaire et logistique face à un nombre élevé de covariables par une approche bayésienne. La sélection de variables est faite en recourant aux méthodes de Monte Carlo par chaînes de Markov, en utilisant l’algorithme de Metropolis-Hastings. On montre qu’en choisissant le processus ponctuel déterminantal comme loi a priori de l’espace des modèles, le sous-ensemble final de variables est meilleur que celui que l’on obtient avec une loi a priori uniforme. / This thesis presents the determinantal point process, a probabilistic model that captures repulsion between points of a certain space. This repulsion is encompassed by a similarity matrix, the kernel matrix, which selects which points are more similar and then less likely to appear in the same subset. This point process gives more weight to subsets characterized by a larger diversity of its elements, which is not the case with the traditional uniform random sampling. Diversity has become a key concept in domains such as medicine, sociology, forensic sciences and behavioral sciences. The determinantal point process is considered a promising alternative to traditional sampling methods, since it takes into account the diversity of selected elements. It is already actively used in machine learning as a subset selection method. Its application in statistics is illustrated with three papers. The first paper presents the consensus clustering, which consists in running a clustering algorithm on the same data, a large number of times. To sample the initials points of the algorithm, we propose the determinantal point process as a sampling method instead of a uniform random sampling and show that the former option produces better clustering results. The second paper extends the methodology developed in the first paper to large-data. Such datasets impose a computational burden since sampling with the determinantal point process is based on the spectral decomposition of the large kernel matrix. We introduce two methods to deal with this issue. These methods also produce better clustering results than consensus clustering based on a uniform sampling of initial points. The third paper addresses the problem of variable selection for the linear model and the logistic regression, when the number of predictors is large. A Bayesian approach is adopted, using Markov Chain Monte Carlo methods with Metropolis-Hasting algorithm. We show that setting the determinantal point process as the prior distribution for the model space selects a better final model than the model selected by a uniform prior on the model space.