81 |
Contribution to the analysis and understanting of electrical-grid signals with signal processing and machine learning techniques / Contribution à l'analyse et à la compréhension des signaux des réseaux électriques par des techniques issues du traitement du signal et de l'apprentissage machineNguyen, Thien-Minh 20 September 2017 (has links)
Ce travail de thèse propose des approches d’identification et de reconnaissance des harmoniques de courant qui sont basées sur des stratégies d’apprentissage automatique. Les approches proposées s’appliquent directement dans les dispositifs d’amélioration de la qualité de l’énergie électrique.Des structures neuronales complètes, dotées de capacités d’apprentissage automatique, ont été développées pour identifier les composantes harmoniques d’un signal sinusoïdal au sens large et plus spécifiquement d’un courant alternatif perturbé par des charges non linéaires. L’identification des harmoniques a été réalisée avec des réseaux de neurones de type Multi–Layer Perceptron (MLP). Plusieurs schémas d’identification ont été développés, ils sont basés sur un réseau MLP composé de neurones linéaire ou sur plusieurs réseaux MLP avec des apprentissages spécifiques. Les harmoniques d’un signal perturbé sont identifiées avec leur amplitude et leur phase, elles peuvent servir à générer des courants de compensation pour améliorer la forme du courant électrique. D’autres approches neuronales a été développées pour reconnaître les charges. Elles consistent en des réseaux MLP ou SVM (Support Vector Machine) et fonctionnent en tant que classificateurs. Leur apprentissage permet à partir des harmoniques de courant de reconnaître le type de charge non linéaire qui génère des perturbations dans le réseau électrique. Toutes les approches d’identification et de reconnaissance des harmoniques ont été validées par des tests de simulation à l’aide des données expérimentales. Des comparaisons avec d’autres méthodes ont démontré des performances supérieures et une meilleure robustesse. / This thesis proposes identifying approaches and recognition of current harmonics that are based on machine learning strategies. The approaches are applied directly in the quality improvement devices of electric energy and in energy management solutions. Complete neural structures, equipped with automatic learning capabilities have been developed to identify the harmonic components of a sinusoidal signal at large and more specifically an AC disturbed by non–linear loads. The harmonic identification is performed with multilayer perceptron neural networks (MLP). Several identification schemes have been developed. They are based on a MLP neural network composed of linear or multiple MLP networks with specific learning. Harmonics of a disturbed signal are identified with their amplitude and phases. They can be used to generate compensation currents fed back into the network to improve the waveform of the electric current. Neural approaches were developed to distinguish and to recognize the types of harmonics and is nonlinear load types that are at the origin. They consist of MLP or SVM (Support Vector Machine) acting as classifier that learns the harmonic profile of several types of predetermined signals and representative of non–linear loads. They entry are the parameters of current harmonics of the current wave. Learning can recognize the type of nonlinear load that generates disturbances in the power network. All harmonics identification and recognition approaches have been validated by simulation tests or using experimental data. The comparisons with other methods have demonstrated superior characteristics in terms of performance and robustness.
|
82 |
Umělá neuronová síť pro modelování polí uvnitř automobilu / Artificial neural network for modeling electromagnetic fields in a carKostka, Filip January 2014 (has links)
The project deals with artificial neural networks. After designing and debugging the test data set and the training sample set, we created a multilayer perceptron network in the Neural NetworkToolbox (NNT) of Matlab. When creating networks, we used different training algorithms and algorithms improving the generalization of the network. When creating a radial basis network, we did not use the NNT, but a specific source code in Matlab was written. Functionality of neural networks was tested on simple training and testing patterns. Realistic training data were obtained by the simulation of twelve monoconic antennas operating in the frequency range from 2 to 6 GHz. Antennas were located inside a mathematical model of Octavia II. Using CST simulations, electromagnetic fields in a car were obtained. Trained networks are described by regressive characteristics andthe mean square error of training. Algorithms improving generalization are applied on the created and trained networks. The performance of individual networks is mutually compared.
|
83 |
Hluboké neuronové sítě / Deep Neural NetworksHabrnál, Matěj January 2014 (has links)
The thesis addresses the topic of Deep Neural Networks, in particular the methods regar- ding the field of Deep Learning, which is used to initialize the weight and learning process s itself within Deep Neural Networks. The focus is also put to the basic theory of the classical Neural Networks, which is important to comprehensive understanding of the issue. The aim of this work is to determine the optimal set of optional parameters of the algori- thms on various complexity levels of image recognition tasks through experimenting with created application applying Deep Neural Networks. Furthermore, evaluation and analysis of the results and lessons learned from the experimentation with classical and Deep Neural Networks are integrated in the thesis.
|
84 |
Rekurentní neuronové sítě v počítačovém vidění / Recurrent Neural Networks in Computer VisionKřepský, Jan January 2011 (has links)
The thesis concentrates on using recurrent neural networks in computer vision. The theoretical part describes the basic knowledge about artificial neural networks with focus on a recurrent architecture. There are presented some of possible applications of the recurrent neural networks which could be used for a solution of real problems. The practical part concentrates on face recognition from an image sequence using the Elman simple recurrent network. For training there are used the backpropagation and backpropagation through time algorithms.
|
85 |
Training a Multilayer Perceptron to predict the final selling price of an apartment in co-operative housing society sold in Stockholm city with features stemming from open data / Träning av en “Multilayer Perceptron” att förutsäga försäljningspriset för en bostadsrättslägenhet till försäljning i Stockholm city med egenskaper från öppna datakällorTibell, Rasmus January 2014 (has links)
The need for a robust model for predicting the value of condominiums and houses are becoming more apparent as further evidence of systematic errors in existing models are presented. Traditional valuation methods fail to produce good predictions of condominium sales prices and systematic patterns in the errors linked to for example the repeat sales methodology and the hedonic pricing model have been pointed out by papers referenced in this thesis. This inability can lead to monetary problems for individuals and in worst-case economic crises for whole societies. In this master thesis paper we present how a predictive model constructed from a multilayer perceptron can predict the price of a condominium in the centre of Stockholm using objective data from sources publicly available. The value produced by the model is enriched with a predictive interval using the Inductive Conformal Prediction algorithm to give a clear view of the quality of the prediction. In addition, the Multilayer Perceptron is compared with the commonly used Support Vector Regression algorithm to underline the hallmark of neural networks handling of a broad spectrum of features. The features used to construct the Multilayer Perceptron model are gathered from multiple “Open Data” sources and includes data as: 5,990 apartment sales prices from 2011- 2013, interest rates for condominium loans from two major banks, national election results from 2010, geographic information and nineteen local features. Several well-known techniques of improving performance of Multilayer Perceptrons are applied and evaluated. A Genetic Algorithm is deployed to facilitate the process of determine appropriate parameters used by the backpropagation algorithm. Finally, we conclude that the model created as a Multilayer Perceptron using backpropagation can produce good predictions and outperforms the results from the Support Vector Regression models and the studies in the referenced papers. / Behovet av en robust modell för att förutsäga värdet på bostadsrättslägenheter och hus blir allt mer uppenbart alt eftersom ytterligare bevis på systematiska fel i befintliga modeller läggs fram. I artiklar refererade i denna avhandling påvisas systematiska fel i de estimat som görs av metoder som bygger på priser från repetitiv försäljning och hedoniska prismodeller. Detta tillkortakommandet kan leda till monetära problem för individer och i värsta fall ekonomisk kris för hela samhällen. I detta examensarbete påvisar vi att en prediktiv modell konstruerad utifrån en “Multilayer Perceptron” kan estimera priset på en bostadsrättslägenhet i centrala Stockholm baserad på allmänt tillgängligt data (“Öppen Data”). Modellens resultat har utökats med ett prediktivt intervall beräknat utifrån “Inductive Conformal Prediction”- algoritmen som ger en klar bild över estimatets tillförlitlighet. Utöver detta jämförs “Multilayer Perceptron”-algoritmen med en annan vanlig algoritm för maskinlärande, den så kallade “Support Vector Regression” för att påvisa neurala nätverks kvalité och förmåga att hantera dataset med många variabler. De variabler som används för att konstruera “Multilayer Perceptron”-modellen är sammanställda utifrån allmänt tillgängliga öppna datakällor och innehåller information så som: priser från 5990 sålda lägenheter under perioden 2011- 2013, ränteläget för bostadsrättslån från två av de stora bankerna, valresultat från riksdagsvalet 2010, geografisk information och nitton lokala särdrag. Ett flertal välkända förbättringar för “Multilayer Perceptron”-algoritmen har applicerats och evaluerats. En genetisk algoritm har använts för att stödja processen att hitta lämpliga parametrar till “Backpropagation”-algoritmen. I detta arbete drar vi slutsatsen att modellen kan producera goda förutsägelser med en modell konstruerad utifrån ett neuralt nätverk av typen “Multilayer Perceptron” beräknad med “backpropagation”, och därmed utklassar de resultat som levereras av Support Vector Regression modellen och de studier som refererats i denna avhandling
|
86 |
Skip connection in a MLP network for Parkinson’s classificationSteinholtz, Tim January 2021 (has links)
In this thesis, two different architecture designs of a Multi-Layer Perceptron network have been implemented. One architecture being an ordinary MLP, and in the other adding DenseNet inspired skip connections to an MLP architecture. The models were used and evaluated on the classification task, where the goal was to classify if subjects were diagnosed with Parkinson’s disease or not based on vocal features. The models were trained on an openly available dataset for Parkinson’s classification and evaluated on a hold-out set from this dataset and on two datasets recorded in another sound recording environment than the training data. The thesis searched for the answer to two questions; How insensitive models for Parkinson’s classification are to the sound recording environment and how the proposed skip connections in an MLP model could help improve performance and generalization capacity. The thesis results show that the sound environment affects the accuracy. Nevertheless, it concludes that one would be able to overcome this with more time and allow for good accuracy when models are exposed to data from a new sound environment than the training data. As for the question, if the skip connections improve accuracy and generalization, the thesis cannot draw any broad conclusions due to the data that were used. The models had, in general, the best performance with shallow networks, and it is with deeper networks that the skip connections are argued to help improve these attributes. However, when evaluating on the data from a different sound recording environment than the training data, the skip connections had the best performance in two out of three tests. / I denna avhandling har två olika arkitektur designer för ett artificiellt flerskikts neuralt nätverk implementerats. En arkitektur som följer konventionen för ett vanlig MLP nätverk, samt en ny arkitektur som introducerar DenseNet inspirerade genvägs kopplingar i MLP nätverk. Modellerna användes och utvärderades för klassificering, vars mål var att urskilja försökspersoner som friska eller diagnostiserade med Parkinsons sjukdom baserat på röst attribut. Modellerna tränades på ett öppet tillgänglig dataset för Parkinsons klassificering och utvärderades på en delmängd av denna data som inte hade använts för träningen, samt två dataset som kommer från en annan ljudinspelnings miljö än datan för träningen. Avhandlingen sökte efter svaret på två frågor; Hur okänsliga modeller för Parkinsons klassificering är för ljudinspelnings miljön och hur de föreslagna genvägs kopplingarna i en MLP-modell kan bidra till att förbättra prestanda och generalisering kapacitet. Resultaten av avhandlingen visar att ljudmiljön påverkar noggrannheten, men drar slutsatsen att med mer tid skulle man troligen kunna övervinna detta och möjliggöra god noggrannhet i nya ljudmiljöer. När det kommer till om genvägs kopplingarna förbättrar noggrannhet och generalisering, är avhandlingen inte i stånd att dra några breda slutsatser på grund av den data som användes. Modellerna hade generellt bästa prestanda med grunda nätverk, och det är i djupare nätverk som genvägs kopplingarna argumenteras för att förbättra dessa egenskaper. Med det sagt, om man bara kollade på resultaten på datan som är ifrån en annan ljudinspelnings miljö så hade genvägs arkitekturen bättre resultat i två av de tre testerna som utfördes.
|
87 |
Evaluating deep learning models for electricity spot price forecastingZdybek, Mia January 2021 (has links)
Electricity spot prices are difficult to predict since they depend on different unstable and erratic parameters, and also due to the fact that electricity is a commodity that cannot be stored efficiently. This results in a volatile, highly fluctuating behavior of the prices, with many peaks. Machine learning algorithms have outperformed traditional methods in various areas due to their ability to learn complex patterns. In the last decade, deep learning approaches have been introduced in electricity spot price prediction problems, often exceeding their predecessors. In this thesis, several deep learning models were built and evaluated for their ability to predict the spot prices 10-days ahead. Several conclusions were made. Firstly, it was concluded that rather simple neural network architectures can predict prices with high accuracy, except for the most extreme sudden peaks. Secondly, all the deep networks outperformed the benchmark statistical model. Lastly, the proposed LSTM and CNN provided forecasts which were statistically, significantly superior and had the lowest errors, suggesting they are the most suitable for the prediction task. / Elspotspriser är svåra att förutsäga eftersom de beror på olika instabila och oregelbundna faktorer, och också på grund av att elektricitet är en vara som inte kan lagras effektivt. Detta leder till ett volatilt, fluktuerande beteende hos priserna, med många plötsliga toppar. Maskininlärningsalgoritmer har överträffat traditionella metoder inom olika områden på grund av deras förmåga att lära sig komplexa mönster. Under det senaste decenniet har djupinlärningsmetoder introducerats till problem inom elprisprognostisering och ofta visat sig överlägsna sina föregångare. I denna avhandling konstruerades och utvärderades flera djupinlärningsmodeller på deras förmåga att förutsäga spotpriserna 10 dagar framåt. Den första slutsatsen är att relativt simpla nätverksarkitekturer kan förutsäga priser med hög noggrannhet, förutom för fallen med de mest extrema, plötsliga topparna. Vidare, så övertränade alla djupa neurala nätverken den statistiska modellen som användes som riktmärke. Slutligen, så gav de föreslagna LSTM- och CNN-modellerna prognoser som var statistiskt, signifikant överlägsna de andra och hade de lägsta felen, vilket tyder på att de är bäst lämpade för prognostiseringsuppgiften.
|
88 |
Aerodynamic coefficient prediction using neural networks.Mailema Celestino dos Santos 04 July 2008 (has links)
The present work discusses the application of neural networks for the accurate prediction of aerodynamic coefficients of airfoil and wing-body configurations. Meta-models based on neural-network are able to handle non-linear problems with a large amount of variables. In this highlight, an efficient methodology employing neural networks for predicting aerodynamic coefficients of generic aircraft was developed. Basic aerodynamic coefficients are modeled depending on angle of attack, number of Mach, Reynolds number, and the lift coefficient of the configuration. A database is provided for the neural network, which is initially trained to learn an overall non-linear model dependent on a large number of variables. A new set of data, which can be relatively sparse, is then supplied to the network to produce a new model consistent with the previous model and the new data. The new model is able to accurately estimate in the sparse test data points and thus the obtaining of a result for a generic configuration is relatively an easy and quick task. Because of this, the methodology is highly suited to be incorporated into a multi-disciplinary design and optimization framework, which make extensively use of aerodynamic calculation for using in other applications, to evaluate performance and loads, besides other core tasks. A Multilayer Perceptrons (MLP) network was designed and employed for predicting drag polar curves of generic airfoils for a given Mach and Reynolds number variation. Airfoil geometry is modeled by polynomial functions described by twelve variables.
|
89 |
Comparative analysis of XGBoost, MLP and LSTM techniques for the problem of predicting fire brigade Iiterventions /Cerna Ñahuis, Selene Leya January 2019 (has links)
Orientador: Anna Diva Plasencia Lotufo / Abstract: Many environmental, economic and societal factors are leading fire brigades to be increasingly solicited, and, as a result, they face an ever-increasing number of interventions, most of the time on constant resource. On the other hand, these interventions are directly related to human activity, which itself is predictable: swimming pool drownings occur in summer while road accidents due to ice storms occur in winter. One solution to improve the response of firefighters on constant resource is therefore to predict their workload, i.e., their number of interventions per hour, based on explanatory variables conditioning human activity. The present work aims to develop three models that are compared to determine if they can predict the firefighters' response load in a reasonable way. The tools chosen are the most representative from their respective categories in Machine Learning, such as XGBoost having as core a decision tree, a classic method such as Multi-Layer Perceptron and a more advanced algorithm like Long Short-Term Memory both with neurons as a base. The entire process is detailed, from data collection to obtaining the predictions. The results obtained prove a reasonable quality prediction that can be improved by data science techniques such as feature selection and adjustment of hyperparameters. / Resumo: Muitos fatores ambientais, econômicos e sociais estão levando as brigadas de incêndio a serem cada vez mais solicitadas e, como consequência, enfrentam um número cada vez maior de intervenções, na maioria das vezes com recursos constantes. Por outro lado, essas intervenções estão diretamente relacionadas à atividade humana, o que é previsível: os afogamentos em piscina ocorrem no verão, enquanto os acidentes de tráfego, devido a tempestades de gelo, ocorrem no inverno. Uma solução para melhorar a resposta dos bombeiros com recursos constantes é prever sua carga de trabalho, isto é, seu número de intervenções por hora, com base em variáveis explicativas que condicionam a atividade humana. O presente trabalho visa desenvolver três modelos que são comparados para determinar se eles podem prever a carga de respostas dos bombeiros de uma maneira razoável. As ferramentas escolhidas são as mais representativas de suas respectivas categorias em Machine Learning, como o XGBoost que tem como núcleo uma árvore de decisão, um método clássico como o Multi-Layer Perceptron e um algoritmo mais avançado como Long Short-Term Memory ambos com neurônios como base. Todo o processo é detalhado, desde a coleta de dados até a obtenção de previsões. Os resultados obtidos demonstram uma previsão de qualidade razoável que pode ser melhorada por técnicas de ciência de dados, como seleção de características e ajuste de hiperparâmetros. / Mestre
|
90 |
Preprocessing perceptronsKallin Westin, Lena January 2004 (has links)
Reliable results are crucial when working with medical decision support systems. A decision support system should be reliable but also be interpretable, i.e. able to show how it has inferred its conclusions. In this thesis, the preprocessing perceptron is presented as a simple but effective and efficient analysis method to consider when creating medical decision support systems. The preprocessing perceptron has the simplicity of a perceptron combined with a performance comparable to the multi-layer perceptron. The research in this thesis has been conducted within the fields of medical informatics and intelligent computing. The original idea of the production line as a tool for a domain expert to extract information, build decision support systems and integrate them in the existing system is described. In the introductory part of the thesis, an introduction to feed-forward neural networks and fuzzy logic is given as a background to work with the preprocessing perceptron. Input to a decision support system is crucial and it is described how to gather a data set, decide how many and what kind of inputs to use. Outliers, errors and missing data are covered as well as normalising of the input. Training is done in a backpropagation-like manner where the division of the data set into a training and a test set can be done in several different ways just as the training itself can have variations. Three major groups of methods to estimate the discriminance effect of the preprocessing perceptron are described and a discussion of the trade-off between complexity and approximation strength are included. Five papers are presented in this thesis. Case studies are shown where the preprocessing perceptron is compared to multi-layer perceptrons, statistical approaches and other mathematical models. The model is extended to a generalised preprocessing perceptron and the performance of this new model is compared to the traditional feed-forward neural networks. Results concerning the preprocessing layer and its connection to multivariate decision limits are included. The well-known ROC curve is described and introduced fully into the field of computer science as well as the improved curve, the QROC curve. Finally a tutorial to the program trainGPP is presented. It describes how to work with the preprocessing perceptron from the moment when a data file is provided to the moment when a new decision support system is built.
|
Page generated in 0.078 seconds