A tarefa de encontrar faces em imagens é extremamente complexa, pois pode ocorrer variação de luminosidade, fundos extremamente complexos e objetos que podem se sobrepor parcialmente à face que será localizada, entre outros problemas. Com o avanço na área de visão computacional técnicas mais recentes de processamento de imagens e inteligência artificial têm sido combinadas para desenvolver algoritmos mais eficientes para a tarefa de detecção de faces. Este trabalho apresenta uma metodologia de visão computacional que utiliza redes neurais MLP (Perceptron Multicamadas) para segmentar a cor da pele e a textura da face, de outros objetos presentes em uma imagem de fundo complexo. A imagem resultante é dividida em regiões, e para cada região são extraídas características que são aplicadas em outra rede neural MLP para identificar se naquela região contem face ou não. Para avaliação do software implementado foram utilizados dois banco de imagens, um com imagens padronizadas (Banco AR) e outro banco com imagens adquiridas na Internet contendo faces com diferentes tons de pele e fundo complexo. Os resultados finais obtidos foram de 83% de faces detectadas para o banco de imagens da Internet e 88% para o Banco AR, evidenciando melhores resultados para as imagens deste banco, pelo fato de serem padronizadas, não conterem faces inclinadas e fundo complexo. A etapa de segmentação apesar de reduzir a quantidade de informação a ser processada para os demais módulos foi a que contribuiu para o maior número de falsos negativos. / The task of finding faces in images is extremely complex, as there is variation in brightness, backgrounds and highly complex objects that may overlap partially in the face to be found, among other problems. With the advancement in the field of computer vision techniques latest image processing and artificial intelligence have been combined to develop more efficient algorithms for the task of face detection. This work presents a methodology for computer vision using neural networks MLP (Multilayer Perceptron) to segment the skin color and texture of the face, from other objects present in a complex background image. The resulting image is divided into regions and from each region are extracted features that are applied in other MLP neural network to identify whether this region contains the face or not. To evaluate the software two sets of images were used, images with a standard database (AR) and another database with images acquired from the Internet, containing faces with different skin tones and complex background. The final results were 83% of faces detected in the internet database of images and 88% for the database AR. These better results for the database AR is due to the fact that they are standardized, are not rotated and do not contain complex background. The segmentation step, despite reducing the amount of information being processed for the other modules contributed to the higher number of false negatives.

Cloud computing has emerged to change the way computing is offered and used. Instead of having all the necessary hardware and software to manipulate and to store their data, users just need a mechanism to access the Internet. So, the efficient provisioning on demand of computational resources is a challenge to comply with the needs of users. Thus, there is a problem related to the lack of an underlying mechanism to assist a cloud management system to maintain acceptable levels of Quality of Service (QoS) pro-actively. In this context, this work makes a comparative analysis of the predictive ability of different statistical models in seeking to define the most suitable for resource provisioning in a cloud environment. In this way, linear time series techniques namely ARIMA and ARMAX and nonlinear ones based on neural networks so-called MLP and NARX were applied on a dataset of a cluster from Google. The prediction results of usage of cpu, disk and memory shown that the NARX neural network can predict with low error the expected values, being feasible for application in a provisioning mechanism of servers in cloud computing environments. / A computação em nuvem surgiu para mudar a forma como a computação é oferecida e utilizada. Ao invés de possuir todo o hardware e software necessários para manipular e armazenar seus dados, os usuários apenas necessitam de um mecanismo que acesse a Internet. Com isso, o provisionamento eficiente de recursos computacionais sob demanda é um desafio para atender as necessidades dos usuários. Dessa forma, percebe-se que existe um problema relacionado à necessidade de mecanismos que auxiliem um sistema de gerenciamento de nuvem a manter níveis adequados de qualidade de serviço (QoS) de forma pro-ativa. Nesse contexto, este trabalho faz uma análise comparativa da capacidade de predição de diferentes modelos estatísticos com vistas a definir o mais adequado ao provisionamento de recursos em um ambiente de nuvem. Para isso, foram aplicadas técnicas de séries temporais lineares ARIMA e ARMAX e não lineares baseadas em redes neurais MLP e NARX em um dataset de um cluster de computadores da Google. Os resultados de predição de uso de cpu, memória e disco demonstraram que a rede neural NARX consegue predizer com baixo erro os valores esperados, sendo viável a sua aplicação em um mecanismo de provisionamento de servidores em ambientes de nuvem computacional

Background Applying the Multi-Level Perspective (MLP) on socio-technical transitions, paired with the interdisciplinary framing approach, this paper investigates how incumbent actors of automobility in Germany framed the issue of a 'transition of mobility and transport' ('Verkehrs/Mobilitätswende') in their public communication in 2020. We first identified representatives of industry, science, policy, and media, since the Verkehrs/Mobilitätswende and its implementation measures are contested among these actors. Employing qualitative content analysis, we then screened 325 public documents according to the elements of the framing approach problem definition, causal interpretation, moral evaluation, and treatment recommendation. Results Findings show that most of the actors frame a transformation of transport and mobility as a necessity. Their arguments encompass environmental and climate-related issues as well as infrastructural problems for bikes and public transport caused by the hegemony of automobility. The actors propose a variety of solutions, primarily focusing on technical innovations for cars or on the expansion of different infrastructures to achieve a modal shift towards sustainable mobility. Conclusion This paper demonstrates that there is no common understanding of the problems and solutions to foster a mobility transition, as the diversity of problems and solutions proposed within the frame elements is high and complicates the prevailing implementation gap of the mobility transition. Therefore, MLP should be conceptually and methodologically bridged with the interdisciplinary framing approach, particularly with regard to the transition of mobility and transport.

Pinpointing the source of a software failure based on log files can be a time consuming process. Automated log analysis tools are meant to streamline such processes, and can be used for tasks like failure diagnosis. This thesis evaluates three supervised models for failure diagnosis of clustered log data. The goal of the thesis is to compare the performance of the models on industry data, as a way to investigate whether the chosen ML techniques are suitable in the context of automated log analysis. A Random Forest, an SVM and an MLP are generated from a dataset of 194 failed executions of tests on microservices, that each resulted in a large collection of logs. The models are tuned with random search and compared in terms of precision, recall, F1-score, hold-out accuracy and 5-fold cross-validation accuracy. The hold-out accuracy is calculated as a mean from 50 hold-out data splits, and the cross-validation accuracy is computed separately from a single set of folds. The results show that the Random Forest scores highest in terms of mean hold-out accuracy (90%), compared to the SVM (86%) and the Neural Network (85%). The mean cross-validation accuracy is the highest for the SVM (95%), closely followed by the Random Forest (94%), and lastly the Neural Network (85%). The precision, recall and F1-score are stable and consistent with the hold-out results, although the precision results are slightly higher than the other two measures. According to this evaluation, the Random Forest has the overall highest performance on the dataset when considering the hold-out- and cross-validation accuracies, and also the fact that it has the lowest complexity and thus the shortest training time, compared to the other considered solutions. All in all, the results of the thesis demonstrate that supervised learning is a promising approach to automatize log analysis. / Att identifiera orsaken till en misslyckad mjukvaruexekvering utifrån logg-filer kan vara en tidskrävande process. Verktyg för automatiserad logg-analysis är tänkta att effektivisera sådana processer, och kan bland annat användas för feldiagnos. Denna avhandling tillhandahåller tre övervakade modeller för feldiagnos av klustrad logg-data. Målet med avhandlingen är att jämföra modellernas prestanda på data från näringslivet, i syfte att utforska huruvida de valda maskininlärningsteknikerna är lämpliga för automatiserad logg-analys. En Random Forest, en SVM och en MLP genereras utifrån ett dataset bestående av 194 misslyckade exekveringar av tester på mikrotjänster, där varje exekvering resulterade i en stor uppsättning loggar. Modellerna finjusteras med hjälp av slumpmässig sökning och jämförs via precision, träffsäkerhet, F-poäng, noggrannhet och 5-faldig korsvalidering. Noggrannheten beräknas som medelvärdet av 50 datauppdelningar, och korsvalideringen tas fram separat från en enstaka uppsättning vikningar. Resultaten visar att Random Forest har högst medelvärde i noggrannhet (90%), jämfört med SVM (86%) och Neurala Nätverket (85%). Medelvärdet i korsvalidering är högst för SVM (95%), tätt följt av Random Forest (94%), och till sist, Neurala Nätverket (85%). Precisionen, träffsäkerheten och F-poängen är stabila och i enlighet med noggrannheten, även om precisionen är något högre än de andra två måtten. Enligt den här analysen har Random Forest överlag högst prestanda på datasetet, med hänsyn till noggrannheten och korsvalideringen, samt faktumet att denna modell har lägst komplexitet och därmed kortast träningstid, jämfört med de andra undersökta lösningarna. Sammantaget visar resultaten från denna avhandling att övervakad inlärning är ett lovande tillvägagångssätt för att automatisera logg-analys.

Sentimentanalys är en teknik som syftar till att automatiskt identifiera den känslomässiga tonen i text. Vanligtvis klassificeras texten som positiv, neutral eller negativ. Nackdelen med denna indelning är att nyanser går förlorade när texten endast klassificeras i tre kategorier. En vidareutveckling av denna klassificering är att inkludera ytterligare två kategorier: mycket positiv och mycket negativ. Utmaningen med denna femklassificering är att det blir svårare att uppnå hög träffsäkerhet på grund av det ökade antalet kategorier. Detta har lett till behovet av att utforska olika metoder för att lösa problemet. Syftet med studien är därför att utvärdera olika klassificerare, såsom MLP, CNN och Bi-GRU i kombination med word2vec för att klassificera sentiment i text i fem kategorier. Studien syftar också till att utforska vilken förbehandling som ger högre träffsäkerhet för word2vec.   Utvecklingen av modellerna gjordes med hjälp av SST-datasetet, som är en känd dataset inom finfördelad sentimentanalys. För att avgöra vilken förbehandling som ger högre träffsäkerhet för word2vec, förbehandlades datasetet på fyra olika sätt. Dessa innefattar enkel förbehandling (EF), samt kombinationer av vanliga förbehandlingar som att ta bort stoppord (EF+Utan Stoppord) och lemmatisering (EF+Lemmatisering), samt en kombination av båda (EF+Utan Stoppord/Lemmatisering). Dropout användes för att hjälpa modellerna att generalisera bättre, och träningen reglerades med early stopp-teknik. För att utvärdera vilken klassificerare som ger högre träffsäkerhet, användes förbehandlingsmetoden som hade högst träffsäkerhet som identifierades, och de optimala hyperparametrarna utforskades. Måtten som användes i studien för att utvärdera träffsäkerheten är noggrannhet och F1-score.   Resultaten från studien visade att EF-metoden presterade bäst i jämförelse med de andra förbehandlingsmetoderna som utforskades. Den modell som hade högst noggrannhet och F1-score i studien var Bi-GRU. / Sentiment analysis is a technique aimed at automatically identifying the emotional tone in text. Typically, text is classified as positive, neutral, or negative. The downside of this classification is that nuances are lost when text is categorized into only three categories. An advancement of this classification is to include two additional categories: very positive and very negative. The challenge with this five-class classification is that achieving high performance becomes more difficult due to the increased number of categories. This has led to the need to explore different methods to solve the problem. Therefore, the purpose of the study is to evaluate various classifiers, such as MLP, CNN, and Bi-GRU in combination with word2vec, to classify sentiment in text into five categories. The study also aims to explore which preprocessing method yields higher performance for word2vec.   The development of the models was done using the SST dataset, which is a well-known dataset in fine-grained sentiment analysis. To determine which preprocessing method yields higher performance for word2vec, the dataset was preprocessed in four different ways. These include simple preprocessing (EF), as well as combinations of common preprocessing techniques such as removing stop words (EF+Without Stopwords) and lemmatization (EF+Lemmatization), as well as a combination of both (EF+Without Stopwords/Lemmatization). Dropout was used to help the models generalize better, and training was regulated with early stopping technique. To evaluate which classifier yields higher performance, the preprocessing method with the highest performance was used, and the optimal hyperparameters were explored. The metrics used in the study to evaluate performance are accuracy and F1-score.   The results of the study showed that the EF method performed best compared to the other preprocessing methods explored. The model with the highest accuracy and F1-score in the study was Bi-GRU.

When it comes to the task of classification the data used for training is the most crucial part. It follows that how this data is processed and presented for the classifier plays an equally important role. This thesis attempts to investigate the performance of multiple classifiers depending on the features that are used, the type of classes to classify and the optimization of said classifiers. The classifiers of interest are support-vector machines (SMO) and multilayer perceptron (MLP), the features tested are word vector spaces and text complexity measures, along with principal component analysis on the complexity measures. The features are created based on the Stockholm-Umeå-Corpus (SUC) and DigInclude, a dataset containing standard and easy-to-read sentences. For the SUC dataset the classifiers attempted to classify texts into nine different text categories, while for the DigInclude dataset the sentences were classified into either standard or simplified classes. The classification tasks on the DigInclude dataset showed poor performance in all trials. The SUC dataset showed best performance when using SMO in combination with word vector spaces. Comparing the SMO classifier on the text complexity measures when using or not using PCA showed that the performance was largely unchanged between the two, although not using PCA had slightly better performance

Climate change has become a catalyst for global action on greenhouse gas emissions. The United Nations Framework Convention on Climate Change orchestrated the Paris Agreement to propel the international community towards implementing definitive carbon abatement plans. These policy commitments are known as Nationally Determined Contributions. However, as of yet many signatories to the Agreement are struggling to align their mitigation pledge with domestic policies. The energy sector is one of the key industries implicit in this carbon abatement process. New energy policies will need to be radically reoriented towards a low-carbon trajectory. In the literature, this pursuit is classed as a socio-technical transition. The Philippines is severely vulnerable to the risks posed by extreme weather patterns exaggerated by increasing temperatures. The country has actively engaged with the climate change discourse but recent trends demonstrate a reversal in low-carbon energy sector planning. Its status as an emerging economy with high potential GDP growth rates increases the urgency to act now to avoid becoming locked-in to an outdated energy system. A discourse and thematic analysis was conducted on key Philippine government texts concerning future energy policy. The approach enabled an exploration of the mechanisms underlying power sector governance in the context of the Paris Agreement. The multi-level perspective provided a conceptual framework for the findings, and enabled the identification of relationships and antagonism within discourses linked to energy system. This framework breaks down the system into three tiers and facilitates analysis of the interplay between landscape pressures, regime resistance and niche experimentation. The results indicated a disparity between the two government agencies on the necessity of low-carbon sector planning. The Philippine Climate Change Commission correlated the benefits of carbon abatement much more closely with the wider goals of sustainable development. The department of energy meanwhile advocated fossil fuel capacity building to meet economic requirements. Divergence in storylines led to a poor alignment between domestic energy policy and the aims of the Paris Agreement. An appreciation of the barriers to a unified overarching mitigation discourse, will assist in the creation of long-term abatement strategies required by the Paris Agreement.

Drug development is a long and expensive process. It is often not until potential drug candidates are administered to humans that accurate quantification of their pharmacokinetic characteristics is achieved. The goal of developing quantitative structure-pharmacokinetic relationships (QSPkRs) is to relate the molecular structure of a chemical entity with its pharmacokinetic characteristics. In this thesis artificial neural networks (ANNs) were used to construct in silico predictive QSPkRs for various pharmacokinetic parameters using different drug data sets. Drug pharmacokinetic data for all studies were taken from the literature. Information for model construction was extracted from drug molecular structure. Numerous theoretical descriptors were generated from drug structure ranging from simple constitutional and functional group counts to complex 3D quantum chemical numbers. Subsets of descriptors were selected which best modeled the target pharmacokinetic parameter(s). Using manual selective pruning, QSPkRs for physiological clearances, volumes of distribution, and fraction bound to plasma proteins were developed for a series of beta-adrenoceptor antagonists. All optimum ANN models had training and cross-validation correlations close to unity, while testing was performed with an independent set of compounds. In most cases the ANN models developed performed better than other published ANN models for the same drug data set. The ability of ANNs to develop QSPkRs with multiple target outputs was investigated for a series of cephalosporins. Multilayer perceptron ANN models were constructed for prediction of half life, volume of distribution, clearances (whole body and renal), fraction excreted in the urine, and fraction bound to plasma proteins. The optimum model was well able to differentiate compounds in a qualitative manner while quantitative predictions were mostly in agreement with observed literature values. The ability to make simultaneous predictions of important pharmacokinetic properties of a compound made this a valuable model. A radial-basis function ANN was employed to construct a quantitative structure-bioavailability relationship for a large, structurally diverse series of compounds. The optimum model contained descriptors encoding constitutional through to conformation dependent solubility characteristics. Prediction of bioavailability for the independent testing set were generally close to observed values. Furthermore, the optimum model provided a good qualitative tool for differentiating between drugs with either low or high experimental bioavailability. QSPkR models constructed with ANNs were compared with multilinear regression models. ANN models were shown to be more effective at selecting a suitable subset of descriptors to model a given pharmacokinetic parameter. They also gave more accurate predictions than multilinear regression equations. This thesis presents work which supports the use of ANNs in pharmacokinetic modeling. Successful QSPkRs were constructed using different combinations of theoretically-derived descriptors and model optimisation techniques. The results demonstrate that ANNs provide a valuable modeling tool that may be useful in drug discovery and development.

Cette étude traite de l'identification de système dynamique non-linéaire. Une architecture multimodèle capable de surmonter certaines difficultés de l'architecture neuronale de type MLP a été étudiée. L'approche multimodèle consiste à représenter un système complexe par un ensemble de modèles de structures simples à validité limitée dans des zones bien définies. A la place de la structure affine des modèles locaux généralement utilisée, cette étude propose une structure polynômiale plus générale, capable de mieux appréhender les non-linéarités locales, réduisant ainsi le nombre de modèles locaux. L'estimation paramétrique d'une telle architecture multimodèle peut se faire suivant une optimisation linéaire, moins coûteuse en temps de calcul que l'estimation paramétrique utilisée dans une architecture neuronale. L'implantation des multimodèles récurrents, avec un algorithme d'estimation paramétrique plus souple que l'algorithme de rétro-propagation du gradient à travers à travers le temps utilisé pour le MLP récurrent a également été effectuée. Cette architecture multimodèle permet de représenter plus facilement des modèles non-linéaires bouclés tels que les modèles NARMAX et NOE. La détermination du nombre de modèles locaux dans une architecture multimodèle nécessite la décomposition (le partitionnement) de l'espace de fonctionnement du système en plusieurs sous-espaces où sont définies les modèles locaux. Des modes de partitionnement du système en plusieurs sous-espaces où sont définies les modèles locaux. Des modes de partitionnement flou (basé sur les algorithmes de "fuzzy-c-means", de "Gustafson et Kessel" et du "substractive clustering") ont été présentés. L'utilisation de telles méthodes nécessite l'implantation d'une architecture multimodèle où les modèles locaux peuvent être de structures différentes : polynômiales de degrés différents, neuronale ou polynômiale et neuronale. Une architecture multimodèle hétérogène répondant à ses exigences a été proposée, des algorithmes d'identification structurelles et paramétriques ont été présentés. Une étude comparative entre les architectures MLP et multimodèle a été menée. Le principal atout de l'architecture mudltimodèle par rapport à l'architecture neuronale de type MLP est la simplicité de l'estimation paramétrique. Par ailleurs, l'utilisation dans une architecture multimodèle d'un mode de partitionnement basé sur la classification floue permet de déterminer facilement le nombre de modèles locaux, alors que la détermination du nombre de neurones cachés pour une architecture MLP reste une tâche difficile.

Drug development is a long and expensive process. It is often not until potential drug candidates are administered to humans that accurate quantification of their pharmacokinetic characteristics is achieved. The goal of developing quantitative structure-pharmacokinetic relationships (QSPkRs) is to relate the molecular structure of a chemical entity with its pharmacokinetic characteristics. In this thesis artificial neural networks (ANNs) were used to construct in silico predictive QSPkRs for various pharmacokinetic parameters using different drug data sets. Drug pharmacokinetic data for all studies were taken from the literature. Information for model construction was extracted from drug molecular structure. Numerous theoretical descriptors were generated from drug structure ranging from simple constitutional and functional group counts to complex 3D quantum chemical numbers. Subsets of descriptors were selected which best modeled the target pharmacokinetic parameter(s). Using manual selective pruning, QSPkRs for physiological clearances, volumes of distribution, and fraction bound to plasma proteins were developed for a series of beta-adrenoceptor antagonists. All optimum ANN models had training and cross-validation correlations close to unity, while testing was performed with an independent set of compounds. In most cases the ANN models developed performed better than other published ANN models for the same drug data set. The ability of ANNs to develop QSPkRs with multiple target outputs was investigated for a series of cephalosporins. Multilayer perceptron ANN models were constructed for prediction of half life, volume of distribution, clearances (whole body and renal), fraction excreted in the urine, and fraction bound to plasma proteins. The optimum model was well able to differentiate compounds in a qualitative manner while quantitative predictions were mostly in agreement with observed literature values. The ability to make simultaneous predictions of important pharmacokinetic properties of a compound made this a valuable model. A radial-basis function ANN was employed to construct a quantitative structure-bioavailability relationship for a large, structurally diverse series of compounds. The optimum model contained descriptors encoding constitutional through to conformation dependent solubility characteristics. Prediction of bioavailability for the independent testing set were generally close to observed values. Furthermore, the optimum model provided a good qualitative tool for differentiating between drugs with either low or high experimental bioavailability. QSPkR models constructed with ANNs were compared with multilinear regression models. ANN models were shown to be more effective at selecting a suitable subset of descriptors to model a given pharmacokinetic parameter. They also gave more accurate predictions than multilinear regression equations. This thesis presents work which supports the use of ANNs in pharmacokinetic modeling. Successful QSPkRs were constructed using different combinations of theoretically-derived descriptors and model optimisation techniques. The results demonstrate that ANNs provide a valuable modeling tool that may be useful in drug discovery and development.