In a changing global energy market where the decarbonization of the economy and the demand growth are pushing to look for new models away from the existing centralized non-renewable based grid. To do so, households have to take a ‘prosumer’ role; to help them take optimal actions is needed a multi-step ahead forecast of their expected energy production and consumption. In multi-step ahead forecasting there are different strategies to perform the forecast. The single-output: Recursive, Direct, DirRec, and the multi-output: MIMO and DIRMO. This thesis performs a comparison between the performance of the differents strategies in a ‘prosumer’ household; using Artificial Neural Networks, Random Forest and K-Nearest Neighbours Regression to forecast both solar energy production and grid input. The results of this thesis indicates that the methodology proposed performs better than state of the art models in a more detailed household energy consumption dataset. They also indicate that the strategy and model of choice is problem dependent and a strategy selection step should be added to the forecasting methodology. Additionally, the performance of the Recursive strategy is always far from the best while the DIRMO strategy performs similarly. This makes the latter a suitable option for exploratory analysis.

[pt] Este trabalho se propõe a desenvolver uma reflexão sobre o processo de intersubjetividade que se inicia desde os primórdios da relação mãe-bebê. Nosso enfoque visa o estudo da questão da interação entre o eu e o outro em um momento em que o outro se encontra em uma posição fronteiriça, na qual, ao mesmo tempo em que é espelho, semelhante, ainda se mantém outro. Utilizando como base principal a teoria psicanalítica de Winnicott, abordamos o processo de subjetivação ressaltando sua dimensão intersubjetiva criada mutuamente pelo par mãe-bebê. Dessa forma, tanto a constituição psíquica do bebê quanto o tornar-se mãe de um bebê específico, são considerados processos construídos a partir do diálogo não verbal, que se estabelece entre a mãe e o recém-nascido na experiência paradoxal de estar-em-um e estar separado. / [en] This work proposes to develop a reflection about the process of intersubjectivity that begins during the initial relationship between a mother and her baby. Our approach seeks to study the interaction between the Self and the Other at a moment in which the Other finds itself in a conflicting position, in which at the same time it mirrors and is similar while still remains the Other. Using as a principal base the psychoanalytical theory of Winnicott, we approach the process of subjectivation highlighting its intersubjective dimension mutually created by the mother-child pair. In this manner, the psychological development of the baby as well as the becoming a mother of a specific baby are considered processes built through a non-verbal dialogue which is established between the mother and the newborn in the paradoxical experience of being one and being separate.

Cette étude a été effectuée pour améliorer l'efficacité des mesures de stockage de carbone par des techniques de télédétection dans les plantations de Para rubber (Hevea brasiliensis) en Thaïlande. Les estimations des méthodes actuelles de stockage de carbone s’opèrent à l’aide de la classification classique basée sur le système des pixels basée sur des images de moyenne résolution et produit donc des résultats d’une grande incertitude. En revanche, dans cette étude, la méthode utilisée est basée sur des images de très haute résolution provenant du satellite THAICHOTE, associés à des mesures sur le terrain, dans le bassin de Mae num Prasae. L’utilisation de l'objet en fonction des classifications, les plantations cartographiées, leur âge et leur circonférence ont été estimées à partir d'un modèle paramétrique dérivé de données spectrales, de texture et 3D. L'étude propose une information de texture plus utile que l'information spectrale pour capturer l’architecture des arbres du couvert et donc l'âge de la canopée. Un spectrale de Global Environment Monitoring (GEMI) et quatre texturales de Homogeneity, Dissimilarity, Contrast et Variance ont été utilisées dans l'ajustement du modèle (régression R2 = 0,87) pour estimer la circonférence et l'âge des arbres tandis que le Canopy Height Model (CHM) de 3D n’était pas autorisée pour construire l'information de classement d'images. Environ 154 km2 des 232 km2 de la zone étudiée sont couverts par des plantations. La quantité totale de la biomasse et des stocks de carbone s’élève à 2,23 mégatonnes et 0,99 mégatonnes C, respectivement avec une incertitude de 11%. En 2011, la superficie totale séquestrée était de 121 tCO2 par des plantations. / This study explored to the improve efficiency of measurements of carbon stock by remote sensing techniques on Para rubber (Hevea brasiliensis) plantations in East Thailand. Current methods of carbon stock estimations use classical pixel based classification on middle-resolution images and thus produce results with a large uncertainty. In this study, the method use very high resolution images from the THAICHOTE satellite, associated to field measurements to estimates the carbon stock and its evolution in the Mae num Prasae watershed. Using object based classifications, the plantations have been mapped and their age and girth have been estimated from a parametric model derived from spectral, textural, 3D information and field data. The results of this study show that these data can be used to map Para rubber plantation and distinguish age classes of trees in the plantations. The study propose that textural information is more useful than spectral information to capture tree canopy architecture and thus the age of the canopy. One spectral of Global Environment Monitoring (GEMI) and four textural information of Homogeneity, Dissimilarity, Contrast and Variance were used in the fit model (multiple linear regression R2=0.87) for estimating the Para rubber tree girth and age while the 3D information (canopy height model: CHM) was not appropriated to build the image classification information. Around 154 km2 of the 232 km2 of the studied area are covered by Para rubber plantations. The total amount of biomass and carbon stocks are 2.23 Megatons and 0.99 Megatons C respectively with uncertainty of 11%. In 2011, the total area sequestered 121 tCO2 by Para rubber plantations.

Les récepteurs nucléaires des xénobiotiques et les enzymes métaboliques P450 (CYP) constituent les principaux éléments contrôlant la biotransformation des médicaments, ainsi que le maintien de des barrières physiologiques au niveau périphérique, plus particulièrement, dans le foie et dans l’intestin. Plusieurs études ont mis en évidence la présence des CYP ainsi que les récepteurs nucléaires contrôlant leur expression, tels que le Constitutive Androstane Receptor et le Pregnane Xenobiotic Receptor (CAR et PXR). Des résultats précédant indiquent la surexpression des CYP2E1 et CYP3A4 dans des tissus et des cellules isolées du cerveau de patients épileptiques pharmacorésistants. L’importance de ces résultats réside dans le rôle du CYP2E1 et CYP3A4 dans la biotransformation de plusieurs médicaments antiépileptiques (MAE) suggérant ainsi un mécanisme de pharmacorésistance aux médicaments. Contrairement aux autres récepteurs nucléaires, les fonctions physiologiques des récepteurs nucléaires des xénobiotiques au niveau vasculaire sont mal connues. Nos résultats montrent des changements spatio-temporaux de l’expression des CYP2E1 et CYP3A4 dans le cerveau après une crise aigüe et pendant le processus d’épileptogenèse chez la souris. Une exposition in vivo et in vitro au MAE Phénytoïne induit une augmentation du niveau du CYP2E1. Le Phénobarbital et la Carbamazépine n’ont pas eu d’effet. Les souris privées des récepteurs nucléaires des xénobiotiques (PXR KO et CAR KO) ne présentent pas de changement de niveau basal des CYP dans le cerveau. Cependant, les souris CAR KO présentent des dysfonctionnements neuronaux (altération de la mémoire, comportement anxieux et une diminution de l’intensité des rythmes EEG entre 3.5-7 Hz) et des modifications caractérisées par une augmentation de perméabilité vasculaire et une dispersion des neurones hippocampiques. L’ensemble des résultats indique une régulation dynamique des CYP dans le cerveau avec une extension de l’impact des récepteurs des xénobiotiques à des fonctions neurovasculaires. / Xenobiotic nuclear receptors and P450 metabolic enzymes (CYP) are pivotal controllers of drug biotransformation and barrier functions in peripheral organs, including the intestine and the liver. Accumulating evidence suggested that, in human, central nervous system cells express significant levels of P450 enzymes and their upstream regulators e.g. Constitutive Androstane and Pregnane Xenobiotic Receptors (CAR, PXR). We previously showed the increased and ectopic CYP2E1 and CYP3A4 expression in brain specimens or cells obtained from drug resistant epileptic patient. These results are significant as CYP2E1 and CYP3A4 are responsible for the metabolism of several antiepileptic drugs (AED), therefore, suggesting a possible new mechanism of drug resistance. However, the exact determinants of CYP expression in the epileptic brain remain unknown. In addition, the exact role of nuclear xenobiotic receptor in the brain is understudied. The latter represents a significant knowledge gap as nuclear receptors other than the xenobiotic ones were shown to contribute to physiological neurovascular functions. Our results show spatio-temporal changes of CYP2E1 and CYP3A4 brain expression occuring after status epilepticus and during epileptogenesis in mice. Exposure to the AED phenytoin, phenobarbital, but not carbamazepine, increased brain CYP2E1 expression in vivo and in vitro. Lack of the specific xenobiotic receptors CAR and PXR did not impact basal CYP brain levels. However, we found an unexpected contribution of CAR to neuronal dysfunctions (memory impairment, anxiety like-behavior and decrease 3.5-7 Hz EEG waves) and neurovascular development, as indicated by increase vascular permeability and hippocampal neuronal dispersion. These results depict a dynamic regulation of P450 enzymes in the brain also expanding the impact of xenobiotic receptors to previously unexplored neurovascular functions.

Uppsatsen undersöker om valet av hanteringsmetod av ordinalskalevariabler kan kopplas till den linjära regressionsmodellens prediceringsförmåga. Antalet hanteringsmetoder som undersöks begränsas till: dummyvariabler, summerat index och viktat summerat index. Stöd för alla tre metoder finns i litteraturen även om det för indexeringarna finns viss förvirring i begreppsanvändning. Med summerat index menas, i denna uppsats, när flera indikatorer som mätts med likertskalor adderas till en ny variabel som antas mäta den latenta variabeln. Med viktat summerat index tas det hänsyn till att indikatorerna kan vara något överlappande eller ha olika stor betydelse för den latenta variabeln. Därför appliceras någon form av viktningsanalys på indikatorerna innan summering sker, denna uppsats använder principalkomponentsanalys. K-delad korsvalidering har nyttjats som främsta analysverktyg för att kunna jämföra de olika hanteringsmetoderna. Jämförandet sker på basis av fyra jämförande mått: R2 , Steins  R2j , RMSE samt MAE. Resultaten indikerar att modeller med dummyvariabler har bäst prediceringsförmåga men det ska förstås utifrån att modellerna hade problem med att uppfylla den linjära regressionsmetodens antaganden. Alla tre hanteringsmetoder har sina för- och nackdelar och därför behöver valet av hanteringsmetod alltid ske med hänsyn till aktuell undersökning.

[pt] O presente estudo se baseia nas histórias de vida de mulheres/mães usuárias de crack, que tiveram seus filhos(as) numa maternidade pública e, que após avaliação de profissionais, foram consideradas sem condições de cuidarem sozinhas de seus bebês. Por medida de proteção à criança em situação de risco social, os casos sofreram intervenção do Judiciário, alguns deles com a suspensão do poder familiar. O que motivou realizar este estudo foi o desejo de identificar na história de vida dessas mulheres um cenário de múltiplas violações de direitos que lhes acometem muito antes do nascimento do bebê. São histórias atravessadas pela pobreza e por diversos episódios de violência. A drogadição se apresenta como um fator de desagregação familiar e de maus tratos contra a própria mulher. Conforme relatos das próprias mulheres, o interesse pela droga surge, a princípio, por curiosidade e influência de amigos; depois, acontece por necessidade e dependência. Nas entrevistas realizadas com três mulheres/mães foi possível perceber a ausência de uma série de ações políticas capazes de prevenir esse quadro de vulnerabilidade, ratificadas pelo depoimento dos profissionais envolvidos neste estudo. Buscou-se, então, encontrar nos seus depoimentos a presença ou não das ações de cuidado, pelo viés público, capazes de garantir a qualidade de vida de pessoas seguramente consideradas cidadãs por um instrumento legal, como a Constituição Federal de 1988. Nessa discussão, apresenta-se uma análise da proteção social brasileira conjugada com a política sobre as drogas, a fim de auxiliar nas reflexões sobre o esfacelamento do direito a uma vida digna que, posteriormente, torna os indivíduos culpados pelos seus próprios descaminhos. / [en] This study is based on the life stories of women/mothers who are addicted to crack and had their babies delivered at a public hospital. These women, having been assessed by professionals, were considered unfit to raise their children on their own. As a protective measure for the child in risky situation, these cases were mediated by the judiciary. In some cases, the family power was suspended. The motivation for this project was the attempt to identify the life-stories I came across in working behind each one of these women. These is a scenario of multiple violations of civil rights through which they endured long before the child is born. These stories are also drenched in violence and poverty. Drug addiction is presented as one on the agents of family break up and women abuse and, as such, leads up to the impossibility of motherhood. As many of them testify, drug use begins as a consequence of curiosity. Later on it becomes a necessity and addiction. While interviewing three of these women/mothers, we could notice the absence of policies and services capable of preventing this state of vulnerability. All that was confirmed by the professionals involved in this study. Therefore, we tried to find in their testimonies the presence or absence of care they received carethrough public policies domain. Measures that couldassure quality of life for all those who are indeed by law considered citizens by legal instrument, as it is stated in the Federal Constitution of 1988. In this project, we included an analysis of the Brazilian social protection system combined with existing drug policies. Our goal is to try to understand the breakdown in their rights to a life worth-living that consequently find those individuals guilty of their own misfortune. We also discussed the role of a women historically, in a time when they are expected to care and protect. However, whether it is caused by desire, lack of care or protection, the women we refer to in this study cannot perform those duties that are socially designed for her.

Today the process of sorting second-hand clothes and textiles is mostly manual. In this master’s thesis, methods for automating this process as well as improving the manual sorting process have been investigated. The methods explored include the automatic prediction of price and intended usage for second-hand clothes, as well as different types of image retrieval to aid manual sorting. Two models were examined: CLIP, a multi-modal model, and MAE, a self-supervised model. Quantitatively, the results favored CLIP, which outperformed MAE in both image retrieval and prediction. However, MAE may still be useful for some applications in terms of image retrieval as it returns items that look similar, even if they do not necessarily have the same attributes. In contrast, CLIP is better at accurately retrieving garments with as many matching attributes as possible. For price prediction, the best model was CLIP. When fine-tuned on the dataset used, CLIP achieved an F1-Score of 38.08 using three different price categories in the dataset. For predicting the intended usage (either reusing the garment or exporting it to another country) the best model managed to achieve an F1-Score of 59.04.

A key challenge for maximizing the world’s food supply is crop yield prediction. In this study, three machine models are used to predict the fresh weight (yield) of lettuce plants that are grown inside indoor farms hydroponically using the vertical farming infrastructure, namely, support vector regressor (SVR), random forest regressor (RFR), and deep neural network (DNN).The climate data, nutrient data, and plant growth data are passed as input to train the models to understand the growth pattern based on the available features. The study of evaluation metrics majorly covers Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), R-squared, and Adjusted R-squared values.The results of the project have shown that the Random Forest with all the features is the best model having the best results with the least cross-validated MAE score and good cross-validated Adjusted R-squared value considering that the error of the prediction is minimal. This is followed by the DNN model with minor differences in the resulting values. The Support Vector Regressor (SVR) model gave a very poor performance with a huge error value that cannot be afforded in this scenario. In this study, we have also compared various evaluating metrics mentioned above and considered the cross-validated MAE and cross-validated Adjusted R-squared metrics. According to our study, MAE had the lowest error value, which is less sensitive to the outliers and adjusted R-squared value helps to understand the variance of the target variable with the predictor variable and adjust the metric to prevent the issues of overfitting.

[pt] O objetivo deste estudo foi realizar uma pesquisa bibliográfica, principalmente na obra de Winnicott, sobre a temática da amamentação, em sua dimensão intersubjetiva. Foram abordadas as implicações psíquicas do ato de mamar tanto para o bebê quanto para a mãe. As relações iniciais entre a mãe e o bebê se dão a partir da mutualidade, da comunicação sensível, da identificação e da dependência. Estes processos suscitam o encontro, para a mulher em estado de preocupação materna primária, com seu eu arcaico, suas partes bebê, enfatizando que o modo como ela foi cuidada, ou não cuidada, quando bebê, interfere na maneira como ela se relaciona com o seu próprio bebê, no momento em que se torna mãe. Mas o vir a ser mãe também possui influências culturais, as quais atualmente apontam para uma sobrecarga desta posição, donde a exigência que se faz é que a mulher amamente sua cria naturalmente, à demanda livre, o que, para muitas, é altamente amedrontador. Por isso, faz-se necessário que a mãe seja cuidada, principalmente no início da vida do bebê. / [en] The aim of this study was to carry out a survey of the literature, mainly in Winnicott s work, about breastfeeding in its inter-subjective aspect. The psychological implications of the act of breastfeeding were considered, both for the baby and for the mother. Initially, mother-infant relations come about through mutuality, attuned communication, identification and dependence. For the woman in a state of primary maternal preoccupation, these processes give rise to the meeting with her archaic self, her infantile parts, emphasizing that how she was cared for, or not cared for as a baby, interferes in the way she relates to her own baby when she becomes a mother. But the mother-coming-into-being also has cultural influences, currently pointing to an overwhelming position: the mother is to breastfeed her child naturally, on demand, which, for many, is a daunting task. For this reason, it is necessary that the mother is cared for, especially in the beginning of a baby s life.

[pt] Esta dissertação trata da luta das mães nas favelas do Rio de Janeiro em busca de justiça pelos seus filhos assassinados por policias militares e agentes do Estado. O enfrentamento entre as mães e o Estado brasileiro é atravessado por questões de classe, raça e gênero que nos permitem entender de que modo essas mães se relacionam com o Estado. Buscando compreender a luta dessas mulheres, analiso como a população negra, pobre e favelada tem sido historicamente marginalizada pelo Estado brasileiro através da construção de um imaginário da favela como um problema na cidade. Na medida em que as políticas de pacificação das favelas matam seus filhos, as mães saem de suas casas e vão para os espaços públicos trazendo em seus corpos a soberania de terem gerado a vida enquanto que o Estado brasileiro a tira. A luta das mães representa uma disputa por significado de soberania, de poder e saber, pois ao gritarem a dor da perda de seus filhos, elas questionam o caráter soberano do Estado-nação brasileiro. As vozes dessas mães em luta colocam em disputa a ideia de soberania estatal, questão central para as Relações Internacionais. / [en] This dissertation deals with the mothers struggle in favelas of Rio de Janeiro in search of justice for their children murdered by military police and other State agents. The confrontations between mothers and the Brazilian State is crossed by issues of class, race and gender which allow us to understand in which way these mothers are related with the State. Trying to understand the struggle of these women I analyze how the black, poor and slum dwellers population has been historically marginalized by the Brazilian State through the construction of an imaginary of slum as a problem of the city. In as much as the politics of pacification of slums kill their children, the mothers get out of their homes and go to public spaces bringing in their bodies the sovereignty of having given the life meanwhile the Brazilian State take it away. The voices of these mothers in struggle put in dispute the idea of state sovereignty, main issue for International Relations.