Considerando a importÃncia do controle da coagulaÃÃo quÃmica para o processo de tratamento de Ãgua por filtraÃÃo rÃpida, esta dissertaÃÃo propÃe a aplicaÃÃo de redes neurais artificiais para inferÃncia dos sinais de controle de dosagem de coagulantes principal e auxiliar, no processo de coagulaÃÃo quÃmica em uma estaÃÃo de tratamento de Ãgua por filtraÃÃo rÃpida. Para tanto, foi feito uma anÃlise comparativa da aplicaÃÃo de modelos baseados em redes neurais do tipo: alimentada adiante focada atrasada no tempo (FTLFN); alimentada adiante atrasada no tempo distribuÃda (DTLFN); recorrente de Elman (ERN) e auto-regressiva nÃo-linear com entradas exÃgenas (NARX). Da anÃlise comparativa, o modelo baseado em redes NARX apresentou melhores resultados, evidenciando o potencial do modelo para uso em casos reais, o que contribuirà para a viabilizaÃÃo de projetos desta natureza em estaÃÃes de tratamento de Ãgua de pequeno porte. / Considering the importance of the chemical coagulation control for the water treatment by direct filtration, this work proposes the application of artificial neural networks for inference of dosage control signals of principal and auxiliary coagulant, in the chemical coagulation process in a water treatment plant by direct filtration. To that end, was made a comparative analysis of the application of models based on neural networks, such as: Focused Time Lagged Feedforward Network (FTLFN); Distributed Time Lagged Feedforward Network (DTLFN); Elman Recurrent Network (ERN) and Non-linear Autoregressive with exogenous inputs (NARX). From the comparative analysis, the model based on NARX networks showed better results, demonstrating the potential of the model for use in real cases, which will contribute to the viability of projects of this nature in small size water treatment plants.

Image recognition and -classification is becoming more important as the need to be able to process large amounts of images is becoming more common. The aim of this thesis is to compare two types of artificial neural networks, FeedForward Network and Convolutional Neural Network, to see how these compare when performing the task of image recognition. Six models of each type of neural network was created that differed in terms of width, depth and which activation function they used in order to learn. This enabled the experiment to also see if these parameters had any effect on the rate which a network learn and how the network design affected the validation accuracy of the models. The models were implemented using the API Keras, and trained and tested using the dataset CIFAR-10. The results showed that within the scope of this experiment the CNN models were always preferable as they achieved a statistically higher validation accuracy compared to their FFN counterparts.

This project aimed to classified artificiality gener-ated,fake, and authentic,true, telecommunication signals, basedupon their frequency response, using methods from deep learn-ing. Another goal was to accomplish this with the least amountof dimension of data possible. The datasets used contained of anequal amount of measured, provided by Ericsson, and generated,by a WINNER II implementation in Matlab, frequency responses.It was determined that a normalized version of the absolute valueof the complex frequency response was enough information for afeedforward network to do a sufficient classification. To improvethe accuracy of the network we did a hyperparameter search,which allowed us to reach an accuracy of 90 percent on our testdataset. The results show that it is possible for neural networksto differentiate between true and fake telecommunication signalsbased on their frequency response, even if it is hard for a humanto tell the difference. / Målet med det här projektet var att klassificera artificiellt genererade signaler, falska, och riktiga, sanna, telekommunikation signaler med hjälp av signalernas frekvens- svar med djup inlärningsmetoder, deep learning. Ett annat mål med projektet var att klassificera signalerna med minsta möjliga antalet dimensioner av datan. Datasetet som användes bestod av till hälften av uppmät data som Ericsson har tillhandahållit, och till hälften av generad data ifrån en WINNER II modell implementerad i Matlab. En slutsats som kunde dras är att en normaliserad version av beloppet av det komplexa frekvenssvaret innehöll tillräckligt med information för att träna ett feedforward nätverk till att uppnå en hög klassificeringssäkerhet. För att vidare öka tillförlitligheten av nätverket gjordes en hyperparametersökning, detta ökade tillförligheten till 90 procent för testdataseten. Resultaten visar att det är möjligt för neurala nätverk att skilja mellan sanna och falska telekommunikations- signaler baserat på deras frekvenssvar, även om det är svårt för människor att skilja signalerna åt. / Kandidatexjobb i elektroteknik 2020, KTH, Stockholm

In the presented master paper: Analysis of car number plate recognition, optical character recognition (OCR), OCR software, OCR devices and systems are reviewed. Image processing operators and artificial neural networks are presented. Analysis and application of image processing operators for detection of number plate is done. Experimental results of estimation of Kohonen and multilayer feedforward artificial neural network learning parameters are presented. Number plate recognition is performed by the use of multilayer feedforward artificial neural network. Model of number plate recognition system is created. Number plate recognition software works in Microsoft© Windows™ operating system. Software is written with C++ language. Experimental results of system model operation are presented.

The subject of this thesis is neural network acceleration with the goal of reducing the number of floating point multiplications. The theoretical part of the thesis surveys current trends and methods used in the field of neural network acceleration. However, the focus is on the binarization techniques which allow replacing multiplications with logical operators. The theoretical base is put into practice in two ways. First is the GPU implementation of crucial binary operators in the Tensorflow framework with a performance benchmark. Second is an application of these operators in simple image classifier. Results are certainly encouraging. Implemented operators achieve speed-up by a factor of 2.5 when compared to highly optimized cuBLAS operators. The last chapter compares accuracies achieved by binarized models and their full-precision counterparts on various architectures.