To see the other types of publications on this topic, follow the link: Multilayer Perceptron neural network model.
Dissertations / Theses on the topic 'Multilayer Perceptron neural network model'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 43 dissertations / theses for your research on the topic 'Multilayer Perceptron neural network model.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
1
Goosen, Johannes Christiaan. "Comparing generalized additive neural networks with multilayer perceptrons / Johannes Christiaan Goosen." Thesis, North-West University, 2011. http://hdl.handle.net/10394/5552.
Full textAbstract:
In this dissertation, generalized additive neural networks (GANNs) and multilayer perceptrons (MLPs) are studied
and compared as prediction techniques. MLPs are the most widely used type of artificial neural network
(ANN), but are considered black boxes with regard to interpretability. There is currently no simple a priori
method to determine the number of hidden neurons in each of the hidden layers of ANNs. Guidelines exist that
are either heuristic or based on simulations that are derived from limited experiments. A modified version of
the neural network construction with cross–validation samples (N2C2S) algorithm is therefore implemented and
utilized to construct good MLP models. This algorithm enables the comparison with GANN models. GANNs
are a relatively new type of ANN, based on the generalized additive model. The architecture of a GANN is less
complex compared to MLPs and results can be interpreted with a graphical method, called the partial residual
plot. A GANN consists of an input layer where each of the input nodes has its own MLP with one hidden layer.
Originally, GANNs were constructed by interpreting partial residual plots. This method is time consuming and
subjective, which may lead to the creation of suboptimal models. Consequently, an automated construction
algorithm for GANNs was created and implemented in the SAS R
statistical language. This system was called
AutoGANN and is used to create good GANN models.
A number of experiments are conducted on five publicly available data sets to gain insight into the similarities
and differences between GANN and MLP models. The data sets include regression and classification tasks.
In–sample model selection with the SBC model selection criterion and out–of–sample model selection with the
average validation error as model selection criterion are performed. The models created are compared in terms
of predictive accuracy, model complexity, comprehensibility, ease of construction and utility.
The results show that the choice of model is highly dependent on the problem, as no single model always
outperforms the other in terms of predictive accuracy. GANNs may be suggested for problems where interpretability
of the results is important. The time taken to construct good MLP models by the modified N2C2S
algorithm may be shorter than the time to build good GANN models by the automated construction algorithm<br>Thesis (M.Sc. (Computer Science))--North-West University, Potchefstroom Campus, 2011.
2
Wilgenbus, Erich Feodor. "The file fragment classification problem : a combined neural network and linear programming discriminant model approach / Erich Feodor Wilgenbus." Thesis, North-West University, 2013. http://hdl.handle.net/10394/10215.
Full textAbstract:
The increased use of digital media to store legal, as well as illegal data, has created the need
for specialized tools that can monitor, control and even recover this data. An important task
in computer forensics and security is to identify the true le type to which a computer le
or computer le fragment belongs. File type identi cation is traditionally done by means
of metadata, such as le extensions and le header and footer signatures. As a result,
traditional metadata-based le object type identi cation techniques work well in cases where
the required metadata is available and unaltered. However, traditional approaches are not
reliable when the integrity of metadata is not guaranteed or metadata is unavailable. As
an alternative, any pattern in the content of a le object can be used to determine the
associated le type. This is called content-based le object type identi cation.
Supervised learning techniques can be used to infer a le object type classi er by exploiting
some unique pattern that underlies a le type's common le structure. This study builds
on existing literature regarding the use of supervised learning techniques for content-based
le object type identi cation, and explores the combined use of multilayer perceptron neural
network classi ers and linear programming-based discriminant classi ers as a solution to the
multiple class le fragment type identi cation problem.
The purpose of this study was to investigate and compare the use of a single multilayer
perceptron neural network classi er, a single linear programming-based discriminant classi-
er and a combined ensemble of these classi ers in the eld of le type identi cation. The
ability of each individual classi er and the ensemble of these classi ers to accurately predict
the le type to which a le fragment belongs were tested empirically.
The study found that both a multilayer perceptron neural network and a linear programming-
based discriminant classi er (used in a round robin) seemed to perform well in solving
the multiple class le fragment type identi cation problem. The results of combining
multilayer perceptron neural network classi ers and linear programming-based discriminant
classi ers in an ensemble were not better than those of the single optimized classi ers.<br>MSc (Computer Science), North-West University, Potchefstroom Campus, 2013
3
Ridhagen, Markus, and Petter Lind. "A comparative study of Neural Network Forecasting models on the M4 competition data." Thesis, Uppsala universitet, Statistiska institutionen, 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-445568.
Full textAbstract:
The development of machine learning research has provided statistical innovations and further developments within the field of time series analysis. This study seeks to investigate two different approaches on artificial neural network models based on different learning techniques, and answering how well the neural network approach compares with a basic autoregressive approach, as well as how the artificial neural network models compare to each other. The models were compared and analyzed in regards to the univariate forecast accuracy on 20 randomly drawn time series from two different time frequencies from the M4 competition dataset. Forecasting was made dependent on one time lag (t-1) and forecasted three and six steps ahead respectively. The artificial neural network models outperformed the baseline Autoregressive model, showing notably lower mean average percentage error overall. The Multilayered perceptron models performed better than the Long short-term memory model overall, whereas the Long short-term memory model showed improvement on longer prediction time dimensions. As the training were done univariately on a limited set of time steps, it is believed that the one layered-approach gave a good enough approximation on the data, whereas the added layer couldn’t fully utilize its strengths of processing power. Likewise, the Long short-term memory model couldn’t fully demonstrate the advantagements of recurrent learning. Using the same dataset, further studies could be made with another approach to data processing. Implementing an unsupervised approach of clustering the data before analysis, the same models could be tested with multivariate analysis on models trained on multiple time series simultaneously.
4
Gao, Zhenning. "Parallel and Distributed Implementation of A Multilayer Perceptron Neural Network on A Wireless Sensor Network." University of Toledo / OhioLINK, 2014. http://rave.ohiolink.edu/etdc/view?acc_num=toledo1383764269.
Full text
5
Midhall, Ruben, and Amir Parmbäck. "Utvärdering av Multilayer Perceptron modeller för underlagsdetektering." Thesis, Malmö universitet, Fakulteten för teknik och samhälle (TS), 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:mau:diva-43469.
Full textAbstract:
Antalet enheter som är uppkopplade till internet, Internet of Things (IoT), ökar hela tiden. År 2035 beräknas det finnas 1000 miljarder Internet of Things-enheter. Samtidigt som antalet enheter ökar, ökar belastningen på internet-nätverken som enheterna är uppkopplade till. Internet of Things-enheterna som finns i vår omgivning samlar in data som beskriver den fysiska tillvaron och skickas till molnet för beräkning. För att hantera belastningen på internet-nätverket flyttas beräkningarna på datan till IoT-enheten, istället för att skicka datan till molnet. Detta kallas för edge computing. IoT-enheter är ofta resurssnåla enheter med begränsad beräkningskapacitet. Detta innebär att när man designar exempelvis "machine learning"-modeller som ska köras med edge computing måste algoritmerna anpassas utifrån de resurser som finns tillgängliga på enheten. I det här arbetet har vi utvärderat olika multilayer perceptron-modeller för mikrokontrollers utifrån en rad olika experiment. "Machine learning"-modellerna har varit designade att detektera vägunderlag. Målet har varit att identifiera hur olika parametrar påverkar "machine learning"-systemen. Vi har försökt att maximera prestandan och minimera den mängd fysiskt minne som krävs av modellerna. Vi har även behövt förhålla oss till att mikrokontrollern inte haft tillgång till internet. Modellerna har varit ämnade att köras på en mikrokontroller "on the edge". Datainsamlingen skedde med hjälp av en accelerometer integrerad i en mikrokontroller som monterades på en cykel. I studien utvärderas två olika "machine learning"-system, ett som är en kombination av binära klassificerings modeller och ett multiklass klassificerings system som framtogs i ett tidigare arbete. Huvudfokus i arbetet har varit att träna modeller för klassificering av vägunderlag och sedan utvärdera modellerna. Datainsamlingen gjordes med en mikrokontroller utrustad med en accelerometer monterad på en cykel. Ett av systemen lyckas uppnå en träffsäkerhet på 93,1\% för klassificering av 3 vägunderlag. Arbetet undersöker även hur mycket fysiskt minne som krävs av de olika "machine learning"-systemen. Systemen krävde mellan 1,78kB och 5,71kB i fysiskt minne.<br>The number of devices connected to the internet, the Internet of Things (IoT), is constantly increasing. By 2035, it is estimated to be 1,000 billion Internet of Things devices in the world. At the same time as the number of devices increase, the load on the internet networks to which the devices are connected, increases. The Internet of Things devices in our environment collect data that describes our physical environment and is sent to the cloud for computation. To reduce the load on the internet networks, the calculations are done on the IoT devices themselves instead of in the cloud. This way no data needs to be sent over the internet and is called edge computing. In edge computing, however, other challenges arise. IoT devices are often resource-efficient devices with limited computing capacity. This means that when designing, for example, machine learning models that are to be run with edge computing, the models must be designed based on the resources available on the device. In this work, we have evaluated different multilayer perceptron models for microcontrollers based on a number of different experiments. The machine learning models have been designed to detect road surfaces. The goal has been to identify how different parameters affect the machine learning systems. We have tried to maximize the performance and minimize the memory allocation of the models. The models have been designed to run on a microcontroller on the edge. The data was collected using an accelerometer integrated in a microcontroller mounted on a bicycle. The study evaluates two different machine learning systems that were developed in a previous thesis. The main focus of the work has been to create algorithms for detecting road surfaces. The data collection was done with a microcontroller equipped with an accelerometer mounted on a bicycle. One of the systems succeeds in achieving an accuracy of 93.1\% for the classification of 3 road surfaces. The work also evaluates how much physical memory is required by the various machine learning systems. The systems required between 1.78kB and 5,71kB of physical memory.
6
Albarakati, Noor. "FAST NEURAL NETWORK ALGORITHM FOR SOLVING CLASSIFICATION TASKS." VCU Scholars Compass, 2012. http://scholarscompass.vcu.edu/etd/2740.
Full textAbstract:
Classification is one-out-of several applications in the neural network (NN) world. Multilayer perceptron (MLP) is the common neural network architecture which is used for classification tasks. It is famous for its error back propagation (EBP) algorithm, which opened the new way for solving classification problems given a set of empirical data. In the thesis, we performed experiments by using three different NN structures in order to find the best MLP neural network structure for performing the nonlinear classification of multiclass data sets. A developed learning algorithm used here is the batch EBP algorithm which uses all the data as a single batch while updating the NN weights. The batch EBP speeds up training significantly and this is also why the title of the thesis is dubbed 'fast NN …'. In the batch EBP, and when in the output layer a linear neurons are used, one implements the pseudo-inverse algorithm to calculate the output layer weights. In this way one always finds the local minimum of a cost function for a given hidden layer weights. Three different MLP neural network structures have been investigated while solving classification problems having K classes: one model/K output layer neurons, K separate models/One output layer neuron, and K joint models/One output layer neuron. The extensive series of experiments performed within the thesis proved that the best structure for solving multiclass classification problems is a K joint models/One output layer neuron structure.
7
Birkmire, Brian Michael. "Weapon Engagement Zone Maximum Launch Range Approximation using a Multilayer Perceptron." Wright State University / OhioLINK, 2011. http://rave.ohiolink.edu/etdc/view?acc_num=wright1313763379.
Full text
8
Ignatavičienė, Ieva. "Tiesioginio sklidimo neuroninių tinklų sistemų lyginamoji analizė." Master's thesis, Lithuanian Academic Libraries Network (LABT), 2012. http://vddb.laba.lt/obj/LT-eLABa-0001:E.02~2012~D_20120801_133809-03141.
Full textAbstract:
Pagrindinis darbo tikslas – atlikti kelių tiesioginio sklidimo neuroninių tinklų sistemų lyginamąją analizę siekiant įvertinti jų funkcionalumą.
Šiame darbe apžvelgiama: biologinio ir dirbtinio neuronų modeliai, neuroninių tinklų klasifikacija pagal jungimo konstrukciją (tiesioginio sklidimo ir rekurentiniai neuroniniai tinklai), dirbtinių neuroninių tinklų mokymo strategijos (mokymas su mokytoju, mokymas be mokytojo, hibridinis mokymas). Analizuojami pagrindiniai tiesioginio sklidimo neuroninių tinklų metodai: vienasluoksnis perceptronas, daugiasluoksnis perceptronas realizuotas „klaidos skleidimo atgal” algoritmu, radialinių bazinių funkcijų neuroninis tinklas.
Buvo nagrinėjama 14 skirtingų tiesioginio sklidimo neuroninių tinklų sistemos. Programos buvo suklasifikuotos pagal kainą, tiesioginio sklidimo neuroninių tinklo mokymo metodų taikymą, galimybę vartotojui keisti parametrus prieš apmokant tinklą ir techninį programos įvertinimą. Programos buvo įvertintos dešimtbalėje vertinimo sistemoje pagal mokymo metodų įvairumą, parametrų keitimo galimybes, programos stabilumą, kokybę, bei kainos ir kokybės santykį. Aukščiausiu balu įvertinta „Matlab” programa (10 balų), o prasčiausiai – „Sharky NN” (2 balai).
Detalesnei analizei pasirinktos keturios programos („Matlab“, „DTREG“, „PathFinder“, „Cortex“), kurios buvo įvertintos aukščiausiais balais, galėjo apmokyti tiesioginio sklidimo neuroninį tinklą daugiasluoksnio perceptrono metodu ir bent dvi radialinių bazinių funkcijų... [toliau žr. visą tekstą]<br>The main aim – to perform a comparative analysis of several feedforward neural system networks in order to identify its functionality.
The work presents both: biological and artificial neural models, also classification of neural networks, according to connections’ construction (of feedforward and recurrent neural networks), studying strategies of artificial neural networks (with a trainer, without a trainer, hybrid). The main methods of feedforward neural networks: one-layer perceptron, multilayer perceptron, implemented upon “error feedback” algorithm, also a neural network of radial base functions have been considered.
The work has included 14 different feedforward neural system networks, classified according its price, application of study methods of feedforward neural networks, also a customer’s possibility to change parameters before paying for the network and a technical evaluation of a program. The programs have been evaluated from 1 point to 10 points according to the following: variety of training systems, possibility to change parameters, stability, quality and ratio of price and quality. The highest evaluation has been awarded to “Matlab” (10 points), the lowest – to “Sharky NN” (2 points).
Four programs (”Matlab“, “DTREG“, “PathFinder“,”Cortex“) have been selected for a detail analysis. The best evaluated programs have been able to train feedforward neural networks using multilayer perceptron method, also at least two radial base function networks. “Matlab“ and... [to full text]
9
Steinholtz, Tim. "Skip connection in a MLP network for Parkinson’s classification." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-303130.
Full textAbstract:
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.<br>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.
10
Tamas, Wani Théo. "Prévision statistique de la qualité de l’air et d’épisodes de pollution atmosphérique en Corse." Thesis, Corte, 2015. http://www.theses.fr/2015CORT0010/document.
Full textAbstract:
L’objectif de ces travaux de doctorat est de développer un modèle prédictif capable de prévoir correctement les concentrations en polluants du jour pour le lendemain en Corse. Nous nous sommes intéressés aux PM10 et à l’ozone, les deux polluants les plus problématiques sur l’île. Le modèle devait correspondre aux contraintes d’un usage opérationnel au sein d’une petite structure, comme Qualitair Corse, l’association locale de surveillance de la qualité de l’air.La prévision a été réalisée à l’aide de réseaux de neurones artificiels. Ces modèles statistiques offrent une grande précision tout en nécessitant peu de ressources informatiques. Nous avons choisi le Perceptron MultiCouche (PMC), avec en entrée à la fois des mesures de polluants, des mesures météorologiques, et des sorties de modèles de chimie-transport (CHIMERE via la plate-forme AIRES) et de modèles météorologiques (AROME).La configuration des PMC a été optimisée avant leur apprentissage automatique, en conformité avec le principe de parcimonie. Pour en améliorer les performances, une étude de sélection de variables a été au préalable menée. Nous avons comparé l’usage d’algorithmes génétiques, de recuits simulés et d’analyse en composantes principales afin d’optimiser le choix des variables d’entrées. L’élagage du PMC a été également mis en œuvre.Nous avons ensuite proposé un nouveau type de modèle hybride, combinaison d’un classifieur et de plusieurs PMC, chacun spécialisé sur un régime météorologique particulier. Ces modèles, qui demandent un large historique de données d’apprentissage, permettent d’améliorer la prévision des valeurs extrêmes et rares, correspondant aux pics de pollution. La classification non-supervisée a été menée avec des cartes auto-organisatrices couplées à l’algorithme des k-means, ainsi que par classification hiérarchique ascendante. L’analyse de sensibilité à été menée grâce à l’usage de courbes ROC.Afin de gérer les jeux de données utilisés, de mener les expérimentations de manière rigoureuse et de créer les modèles destinés à l’usage opérationnel, nous avons développé l’application « Aria Base », fonctionnant sous Matlab à l’aide de la Neural Network Toolbox.Nous avons également développé l’application « Aria Web » destinée à l’usage quotidien à Qualitair Corse. Elle est capable de mener automatiquement les prévisions par PMC et de synthétiser les différentes informations qui aident la prévision rendues disponibles sur internet par d’autres organismes<br>The objective of this doctoral work is to develop a forecasting model able to correctly predict next day pollutant concentrations in Corsica. We focused on PM10 and ozone, the two most problematic pollutants in the island. The model had to correspond to the constraints of an operational use in a small structure like Qualitair Corse, the local air quality monitoring association.The prediction was performed using artificial neural networks. These statistical models offer a great precision while requiring few computing resources. We chose the MultiLayer Perceptron (MLP), with input data coming from pollutants measurements, meteorological measurements, chemical transport model (CHIMERE via AIRES platform) and numerical weather prediction model (AROME).The configuration of the MLP was optimized prior to machine learning, in accordance with the principle of parsimony. To improve forecasting performances, we led a feature selection study. We compared the use of genetic algorithms, simulated annealing and principal component analysis to optimize the choice of input variables. The pruning of the MLP was also implemented.Then we proposed a new type of hybrid model, combination of a classification model and various MLPs, each specialized on a specific weather pattern. These models, which need large learning datasets, allow an improvement of the forecasting for extreme and rare values, corresponding to pollution peaks. We led unsupervised classification with self organizing maps coupled with k-means algorithm, and with hierarchical ascendant classification. Sensitivity analysis was led with ROC curves.We developed the application “Aria Base” running with Matlab and its Neural Network Toolbox, able to manage our datasets, to lead rigorously the experiments and to create operational models.We also developed the application “Aria Web” to be used daily by Qualitair Corse. It is able to lead automatically the prevision with MLP, and to synthesize forecasting information provided by other organizations and available on the Internet
11
Liberatore, Lorenzo. "Introduction to geometric deep learning and graph neural networks." Bachelor's thesis, Alma Mater Studiorum - Università di Bologna, 2022. http://amslaurea.unibo.it/25339/.
Full textAbstract:
This thesis proposes an introduction to the fundamental concepts of supervised deep learning. Starting from Rosemblatt's Perceptron we will discuss the architectures that, in recent years, have revolutioned the world of deep learning: graph neural networks, which led to the formulation of geometric deep learning. We will then give a simple example of graph neural network, discussing the code that composes it and then test our
architecture on the MNISTSuperpixels dataset, which is a variation of the benchmark dataset MNIST.
12
Rönnholm, Niklas. "A study of limitations and performance in scalable hosting using mobile devices." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-224646.
Full textAbstract:
At present day, distributed computing is a widely used technique, where volunteers support different computing power needs organizations might have. This thesis sought to benchmark distributed computing performance limited to mobile device support since this type of support is seldom done with mobile devices. This thesis proposes two approaches to harnessing computational power and infrastructure of a group of mobile devices. The problems used for benchmarking are small instances of deep learning training. One requirement posed by the mobile devices’ non-static nature was that this should be possible without any significant prior configuration. The protocol used for communication was HTTP. The reason deep-learning was chosen as the benchmarking problem is due to its versatility and variability. The results showed that this technique can be applied successfully to some types of problem instances, and that the two proposed approaches also favour different problem instances. The highest request rate found for the prototype with a 99% response rate was a 2100% increase in efficiency compared to a regular server. This was under the premise that it was provided just below 2000 mobile devices for only particular problem instances.<br>För närvarande är distribuerad databehandling en utbredd teknik, där frivilliga individer stödjer olika organisationers behov av datorkraft. Denna rapport försökte jämföra prestandan för distribuerad databehandling begränsad till enbart stöd av mobila enheter då denna typ av stöd sällan görs med mobila enheter. Rapporten föreslår två sätt att utnyttja beräkningskraft och infrastruktur för en grupp mobila enheter. De problem som används för benchmarking är små exempel på deep-learning. Ett krav som ställdes av mobilenheternas icke-statiska natur var att detta skulle vara möjligt utan några betydande konfigureringar. Protokollet som användes för kommunikation var HTTP. Anledningen till att deeplearning valdes som referensproblem beror på dess mångsidighet och variation. Resultaten visade att denna teknik kan tillämpas framgångsrikt på vissa typer av probleminstanser, och att de två föreslagna tillvägagångssätten också gynnar olika probleminstanser. Den högsta requesthastigheten hittad för prototypen med 99% svarsfrekvens var en 2100% ökning av effektiviteten jämfört med en vanlig server. Detta givet strax under 2000 mobila enheter för vissa speciella probleminstanser.
13
Silva, Aldemário Alves da. "Desenvolvimento e aplicação de Heurística para calcular pesos e bias iniciais para o “Back-Propagation” treinar Rede Neural Perceptron Multicamadas." Universidade Federal Rural do Semi-Árido, 2017. http://bdtd.ufersa.edu.br:80/tede/handle/tede/806.
Full textAbstract:
Submitted by Lara Oliveira (lara@ufersa.edu.br) on 2017-09-08T22:30:39Z
No. of bitstreams: 1
AldemárioAS_DISSERT.pdf: 18856416 bytes, checksum: dcd37bbe9d111ef051c4d27c3481a41f (MD5)<br>Approved for entry into archive by Vanessa Christiane (referencia@ufersa.edu.br) on 2017-09-11T16:27:51Z (GMT) No. of bitstreams: 1
AldemárioAS_DISSERT.pdf: 18856416 bytes, checksum: dcd37bbe9d111ef051c4d27c3481a41f (MD5)<br>Approved for entry into archive by Vanessa Christiane (referencia@ufersa.edu.br) on 2017-09-11T16:28:25Z (GMT) No. of bitstreams: 1
AldemárioAS_DISSERT.pdf: 18856416 bytes, checksum: dcd37bbe9d111ef051c4d27c3481a41f (MD5)<br>Made available in DSpace on 2017-09-11T16:29:16Z (GMT). No. of bitstreams: 1
AldemárioAS_DISSERT.pdf: 18856416 bytes, checksum: dcd37bbe9d111ef051c4d27c3481a41f (MD5)
Previous issue date: 2017-08-18<br>The training of Multilayer Perceptron Neural Network (MLPNN) done by exact algorithm to find the maximum accuracy is NP-hard. Thus, we use the algorithm Back-Propagation who needs a starting point (weights and bias initials) to compute the training of the MLPNN. This research has developed and implemented a heuristic algorithm HeCI - Heuristic to Calculate Weights and Bias Initials - to compute the data to train the MLPNN and return the starting point for the Back-Propagation. HeCI uses Principal Component Analysis, Least Square Method, Probability Density Function of the Normal Gaussian Distribution, two strategic configurations, and partially controls the number of MLPNN training epochs. Experimentally, HeCI was used with Back-Propagation in MLPNN training to recognize patterns and solve data classification problems. Six case studies with datasets between Health, Business and Botany were used in the experiments. The methodology of this research uses Deductive analysis by the Experimental method with Quantitative approach and hypothesis tests: Test of Fridman with post Teste of Tukey HSD Post-hoc and Wilcoxon Test-M W. The results of accuracy have increased significantly improving attested by evaluation of tests of hypotheses, inferringstatistical robustness of the result motivated by HeCI<br>O treinamento de Rede Neural Perceptron Multicamadas (RNPM) feito por algoritmo exato para encontrar a máxima acurácia é NP-Difícil. Sendo assim, usa-se o algoritmo "Back-Propagation" que necessita de um ponto de partida (pesos e bias iniciais) para computar o treinamento da RNPM. Esta pesquisa desenvolveu e aplicou um algoritmo heurístico HeCI - Heurística para Calcular Pesos e Bias Iniciais - para computar os dados de treinamento da RNPM e retornar o ponto de partida para o "Back-Propagation". A HeCI usa Análise de Componentes Principais, Método dos Mínimos Quadrados, Função de Densidade de Probabilidade da Normal Distribuição Gaussiana, duas configurações estratégicas e controla parcialmente o número de épocas de treinamento da RNPM. Experimentalmente, a RNPM foi treinada usando "Back-Propagation" com HeCI, para reconhecer padrões e resolver problemas de classificação de dados. Seis estudos de caso com "datasets" entre as áreas de Saúde, Negócio e Botânica foram usados nos experimentos. A metodologia desta pesquisa usa análise Dedutiva pelo método Experimental com abordagem Quantitativa e testes de hipóteses: Teste de Fridman com Pós Teste de Tukey HSD Post-hoc e Teste de Wilcoxon-M-W. Os resultados de acurácia incrementaram melhoria significativa atestada pela avaliação dos testes de hipóteses, inferindo estatisticamente robustez de resultado motivado pela HeCI<br>2017-09-08
14
Fischer, Manfred M., and Sucharita Gopal. "Neural Network Models and Interregional Telephone Traffic. Comparative Performance Comparisons between Multilayer Feedforward Networks and the Conventional Spatial Interaction Model." WU Vienna University of Economics and Business, 1992. http://epub.wu.ac.at/4206/1/WSG_DP_2792.pdf.
Full text
15
Santos, Rosiane Correia. "LearnInPlanner: uma abordagem de aprendizado supervisionado com redes neurais para solução de problemas de planejamento clássico." Universidade de São Paulo, 2013. http://www.teses.usp.br/teses/disponiveis/100/100131/tde-25012014-115621/.
Full textAbstract:
A busca progressiva no espaço de estados é uma das abordagens mais populares de Planejamento Automatizado. O desempenho dos algoritmos de busca progressiva é influenciado pela heurística independente de domínio utilizada para guiá-lo. Nesse contexto, o foco do presente trabalho consiste em investigar técnicas de aprendizado de máquina supervisionadas que possibilitaram agregar à heurística do plano relaxado, comumente utilizada em abordagens atuais de planejamento, informações sobre o domínio em questão que viessem a ser úteis ao algoritmo de busca. Essas informações foram representadas por meio de um espaço de características do problema de planejamento e uma rede neural MLP foi aplicada para estimar uma nova função heurística para guiar a busca por meio de um processo de regressão não linear. Uma vez que o conjunto de características disponíveis para a construção da nova função heurística é grande, foi necessário a definição de um processo de seleção de características capaz de determinar qual conjunto de características de entrada da rede resultaria em melhor desempenho para o modelo de regressão. Portanto, para a seleção de características, aplicou-se uma abordagem de algoritmos genéticos. Como principal resultado, tem-se uma análise comparativa do desempenho entre a utilização da heurística proposta neste trabalho e a utilização da heurística do plano relaxado para guiar o algoritmo de busca na tarefa de planejamento. Para a análise empírica foram utilizados domínios de diferentes complexidades disponibilizados pela Competições Internacionais de Planejamento. Além dos resultados empíricos e análises comparativas, as contribuições deste trabalho envolvem o desenvolvimento de um novo planejador independente de domínio, denominado LearnInPlanner. Esse planejador utiliza a nova função heurística estimada por meio do processo de aprendizado e o algoritmo de Busca Gulosa para solucionar os problemas de planejamento.<br>The forward state-space search is one of the most popular Automated Planning approaches. The performance of forward search algorithms is affected by the domain-independent heuristic being used. In this context, the focus of this work consisted on investigating techniques of supervised machine learning that make possible to agregate to the relaxed plan heuristic, commonly used in current planning approaches, information about the domain which could be useful to the search algorithm. This information has been represented through a feature space of planning problem and a MLP neural network has been applied to estimate a new heuristic function for guiding the search through a non-linear regression process. Once the set of features available for the construction of the new heuristic function is large, it was necessary to define a feature selection process capable of determining which set of neural network input features would result in the best performance for the regression model. Therefore, for selecting features, an approach of genetic algorithms has been applied. As the main result, one has obtained a comparative performance analysis between the use of heuristic proposed in this work and the use of the relaxed plan heuristic to guide the search algorithm in the planning task. For the empirical analysis were used domains with different complexities provided by the International Planning Competitions. In addition to the empirical results and comparative analysis, the contributions of this work involves the development of a new domain-independent planner, named LearnInPlanner. This planner uses the new heuristic function estimated by the learning process and the Greedy Best-First search algorithm to solve planning problems.
16
Křepský, Jan. "Rekurentní neuronové sítě v počítačovém vidění." Master's thesis, Vysoké učení technické v Brně. Fakulta informačních technologií, 2011. http://www.nusl.cz/ntk/nusl-237029.
Full textAbstract:
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.
17
Kostka, Filip. "Umělá neuronová síť pro modelování polí uvnitř automobilu." Master's thesis, Vysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií, 2014. http://www.nusl.cz/ntk/nusl-220578.
Full textAbstract:
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.
18
Prosperi, Maurizio. "Analysis of the EU MacSharry reform in Italian family farms : a multilayer feed-forward neural network model for evaluating economic and environmental impacts of the direct payments system." Kyoto University, 2005. http://hdl.handle.net/2433/144591.
Full textAbstract:
Kyoto University (京都大学)<br>0048<br>新制・課程博士<br>博士(農学)<br>甲第11837号<br>農博第1527号<br>新制||農||919(附属図書館)<br>学位論文||H17||N4086(農学部図書室)<br>23597<br>UT51-2005-K503<br>京都大学大学院農学研究科生物資源経済学専攻<br>(主査)教授 加賀 爪優, 教授 吉田 昌之, 教授 小田 滋晃<br>学位規則第4条第1項該当
19
BRANDL, ALBERTO. "Techniques for effective virtual sensor development and implementation with application to air data systems." Doctoral thesis, Politecnico di Torino, 2020. http://hdl.handle.net/11583/2842493.
Full text
20
Richter, Jan. "Využití metod umělé inteligence pro simulaci a identifikaci dat v oblasti proudění." Doctoral thesis, Vysoké učení technické v Brně. Fakulta strojního inženýrství, 2019. http://www.nusl.cz/ntk/nusl-391879.
Full textAbstract:
It is possible to simulate an airflow by additives to shoot images and records of such flowing. Additives can be in the form of particles or continuous filaments. A computer evaluation of such data differs depending on the kind of visualization method. This thesis deals with a number of different approaches to determine the airjet shape and airflow velocity in airflow images and records. Exact procedures area sed for these purposes as well as neural networks and genetic algorithms.
21
Muñoz, Mas Rafael. "Multivariate approaches in species distribution modelling: Application to native fish species in Mediterranean Rivers." Doctoral thesis, Universitat Politècnica de València, 2018. http://hdl.handle.net/10251/76168.
Full textAbstract:
This dissertation focused in the comprehensive analysis of the capabilities of some non-tested types of Artificial Neural Networks, specifically: the Probabilistic Neural Networks (PNN) and the Multi-Layer Perceptron (MLP) Ensembles. The analysis of the capabilities of these techniques was performed using the native brown trout (Salmo trutta; Linnaeus, 1758), the bermejuela (Achondrostoma arcasii; Robalo, Almada, Levy & Doadrio, 2006) and the redfin barbel (Barbus haasi; Mertens, 1925) as target species. The analyses focused in the predictive capabilities, the interpretability of the models and the effect of the excess of zeros in the training datasets, which for presence-absence models is directly related to the concept of data prevalence (i.e. proportion of presence instances in the training dataset). Finally, the effect of the spatial scale (i.e. micro-scale or microhabitat scale and meso-scale) in the habitat suitability models and consequently in the e-flow assessment was studied in the last chapter.<br>Esta tesis se centra en el análisis comprensivo de las capacidades de algunos tipos de Red Neuronal Artificial aún no testados: las Redes Neuronales Probabilísticas (PNN) y los Conjuntos de Perceptrones Multicapa (MLP Ensembles). Los análisis sobre las capacidades de estas técnicas se desarrollaron utilizando la trucha común (Salmo trutta; Linnaeus, 1758), la bermejuela (Achondrostoma arcasii; Robalo, Almada, Levy & Doadrio, 2006) y el barbo colirrojo (Barbus haasi; Mertens, 1925) como especies nativas objetivo. Los análisis se centraron en la capacidad de predicción, la interpretabilidad de los modelos y el efecto del exceso de ceros en las bases de datos de entrenamiento, la así llamada prevalencia de los datos (i.e. la proporción de casos de presencia sobre el conjunto total). Finalmente, el efecto de la escala (micro-escala o escala de microhábitat y meso-escala) en los modelos de idoneidad del hábitat y consecuentemente en la evaluación de caudales ambientales se estudió en el último capítulo.<br>Aquesta tesis se centra en l'anàlisi comprensiu de les capacitats d'alguns tipus de Xarxa Neuronal Artificial que encara no han estat testats: les Xarxes Neuronal Probabilístiques (PNN) i els Conjunts de Perceptrons Multicapa (MLP Ensembles). Les anàlisis sobre les capacitats d'aquestes tècniques es varen desenvolupar emprant la truita comuna (Salmo trutta; Linnaeus, 1758), la madrilla roja (Achondrostoma arcasii; Robalo, Almada, Levy & Doadrio, 2006) i el barb cua-roig (Barbus haasi; Mertens, 1925) com a especies objecte d'estudi. Les anàlisi se centraren en la capacitat predictiva, interpretabilitat dels models i en l'efecte de l'excés de zeros a la base de dades d'entrenament, l'anomenada prevalença de les dades (i.e. la proporció de casos de presència sobre el conjunt total). Finalment, l'efecte de la escala (micro-escala o microhàbitat i meso-escala) en els models d'idoneïtat de l'hàbitat i conseqüentment en l'avaluació de cabals ambientals es va estudiar a l'últim capítol.<br>Muñoz Mas, R. (2016). Multivariate approaches in species distribution modelling: Application to native fish species in Mediterranean Rivers [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/76168<br>TESIS
22
Козаченко, А. П. "Інформаційне та програмне забезпечення системи інтелектуального аналізу енергоспоживання будівель". Master's thesis, Сумський державний університет, 2018. http://essuir.sumdu.edu.ua/handle/123456789/72073.
Full textAbstract:
Розроблено алгоритм та програмне забезпечення інтелектуальної системи класифікації будівель за рівнем енергоспоживання на базі комплексної оцінки енерговитрат в зимовий та літній період. В роботі проведено оптимізація параметрів стандартного алгоритму навчання нейромереж зворотного розповсюдження помилки, що дозволило підвищити ефективність сформованого нейромережевого класифікатору. Розроблений алгоритм реалізовано у формі m-сценаріїв середовища для наукових і інженерних розрахунків MATLAB 6.5.
23
Leonavičius, Romas. "Melizmų sintezė dirbtinių neuronų tinklais." Doctoral thesis, Lithuanian Academic Libraries Network (LABT), 2007. http://vddb.library.lt/obj/LT-eLABa-0001:E.02~2006~D_20070112_145929-44906.
Full textAbstract:
Modern methods of speech synthesis are not suitable for restoration of song signals due to lack of vitality and intonation in the resulted sounds. The aim of presented work is to synthesize melismas met in Lithuanian folk songs, by applying Artificial Neural Networks. An analytical survey of rather a widespread literature is presented. First classification and comprehensive discussion of melismas are given. The theory of dynamic systems which will make the basis for studying melismas is presented and finally the relationship for modeling a melisma with nonlinear and dynamic systems is outlined. Investigation of the most widely used Linear Prediction Coding method and possibilities of its improvement. The modification of original Linear Prediction method based on dynamic LPC frame positioning is proposed. On its basis, the new melisma synthesis technique is presented. Developed flexible generalized melisma model, based on two Artificial Neural Networks – a Multilayer Perceptron and Adaline – as well as on two network training algorithms – Levenberg- Marquardt and the Least Squares error minimization – is presented. Moreover, original mathematical models of Fortis, Gruppett, Mordent and Trill are created, fit for synthesizing melismas, and their minimal sizes are proposed. The last chapter concerns experimental investigation, using over 500 melisma records, and corroborates application of the new mathematical models to melisma synthesis of one performer.
24
Predrag, Pecev. "Развој алгоритма и система за дедуктивну предикцију и анализу кретања кошаркашких судија". Phd thesis, Univerzitet u Novom Sadu, Prirodno-matematički fakultet u Novom Sadu, 2017. https://www.cris.uns.ac.rs/record.jsf?recordId=104240&source=NDLTD&language=en.
Full textAbstract:
Докторска дисертација припада области информационих система, са јасним акцентом на употребу неуронских мрежа за решавање проблема вишеструких зависних временских серија који је у овом докторату дефинисан.Основни циљ дисертације је креирање система у форми едукативног софтвера путем којег ће се обучавати младе кошаркашке судијеЈедан од кључих елемената овог доктората јесте симулација хоризонталног видног поља на основу којег се утврђује да ли је резоновано кретање кошаркашких судија било адекватно или није. Стога развијени софтвер поседује споменуту едукативну примену.Како би се реализовао споменути софтвер спроведено је истраживање које је обухватило обучавање великог броја традиционалних вишеслојних перцептрона као и формирање посебне LTR – MDTS структуре неуронске мреже за коју се сматра да је погодна за решавање постојећег проблема. За реализацију симулације хоризонталног видног поља разматрано је више алгоритама из области рачунарске графике а Sweep and Prune алгоритам је парцијално пружио основу за развијени и тренутно имплементирани алгоритам.<br>Doktorska disertacija pripada oblasti informacionih sistema, sa jasnim akcentom na upotrebu neuronskih mreža za rešavanje problema višestrukih zavisnih vremenskih serija koji je u ovom doktoratu definisan.Osnovni cilj disertacije je kreiranje sistema u formi edukativnog softvera putem kojeg će se obučavati mlade košarkaške sudijeJedan od ključih elemenata ovog doktorata jeste simulacija horizontalnog vidnog polja na osnovu kojeg se utvrđuje da li je rezonovano kretanje košarkaških sudija bilo adekvatno ili nije. Stoga razvijeni softver poseduje spomenutu edukativnu primenu.Kako bi se realizovao spomenuti softver sprovedeno je istraživanje koje je obuhvatilo obučavanje velikog broja tradicionalnih višeslojnih perceptrona kao i formiranje posebne LTR – MDTS strukture neuronske mreže za koju se smatra da je pogodna za rešavanje postojećeg problema. Za realizaciju simulacije horizontalnog vidnog polja razmatrano je više algoritama iz oblasti računarske grafike a Sweep and Prune algoritam je parcijalno pružio osnovu za razvijeni i trenutno implementirani algoritam.<br>Doctoral dissertation belongs to the field of information systems, with a clear emphasis on the use of neural networks for solving the problem of multiple dependent time series, which is defined in this doctorate. The main objective of the thesis is to create a system in the form of educational software that will be used druring the training of young basketball referees.One of the key elements of this doctorate is a simulation of a horizontal field of vision on the basis of which it is determined whether the movement of reasoned basketball referees was adequate or not. Therefore developed software has aforementioned educational use. In order to realize the aforementioned software, a research was conducted that included training of a large number of traditional multilayer perceptron neural networks and the formation of special LTR - MDTS neural network structure which is considered to be suitable for solving the presented problem. For the realization of the simulation of the horizontal field of vision a large number of algorithms in the field of computer graphis was considered and Sweep and Prune algorithm partially provided the basis for the developed and currently implemented algorithm.
25
Buhot, Arnaud. "Etude de propriétés d'apprentissage supervisé et non supervisé par des méthodes de Physique Statistique." Phd thesis, Université Joseph Fourier (Grenoble), 1999. http://tel.archives-ouvertes.fr/tel-00001642.
Full textAbstract:
L'objet de cette thèse est l'étude de diverses propriétés d'apprentissage à partir d'exemples par des méthodes de Physique Statistique, notamment, par la méthode des répliques. Des tâches supervisées, correspondant à la classification binaire de données, ainsi que des tâches non supervisées, comme l'estimation paramétrique d'une densité de probabilité, sont considérées. Dans la première partie, une approche variationnelle permet de déterminer la performance de l'apprentissage optimal d'une direction d'anisotropie, et de déduire une fonction de coût permettant d'obtenir ces performances optimales. Dans le cas de l'apprentissage supervisé d'une tâche linéairement séparable, des simulations numériques confirmant nos résultats théoriques ont permis de déterminer les effets de taille finie. Dans le cas d'une densité de probabilité constituée de deux gaussiennes, la performance de l'apprentissage optimal présente de nombreuses transitions de phases en fonction du nombre de données. Ces résultats soulèvent une controverse entre la théorie variationnelle et l'approche bayesienne de l'apprentissage optimal. Dans la deuxième partie, nous étudions deux approches différentes de l'apprentissage de tâches de classification complexes. La première approche considérée est celle des machines à exemples supports. Nous avons étudié une famille de ces machines pour laquelle les séparateurs linéaire et quadratique sont deux cas particuliers. La capacité, les valeurs typiques de la marge et du nombre d'exemples supports, sont déterminées. La deuxième approche considérée est celle d'une machine de parité apprenant avec un algorithme incrémental. Cet algorithme construit progressivement un réseau de neurones à une couche cachée. La capacité théorique obtenue pour l'algorithme considéré est proche de celle de la machine de parité.
26
Narmack, Kirilll. "Dynamic Speed Adaptation for Curves using Machine Learning." Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-233545.
Full textAbstract:
The vehicles of tomorrow will be more sophisticated, intelligent and safe than the vehicles of today. The future is leaning towards fully autonomous vehicles. This degree project provides a data driven solution for a speed adaptation system that can be used to compute a vehicle speed for curves, suitable for the underlying driving style of the driver, road properties and weather conditions. A speed adaptation system for curves aims to compute a vehicle speed suitable for curves that can be used in Advanced Driver Assistance Systems (ADAS) or in Autonomous Driving (AD) applications. This degree project was carried out at Volvo Car Corporation. Literature in the field of speed adaptation systems and factors affecting the vehicle speed in curves was reviewed. Naturalistic driving data was both collected by driving and extracted from Volvo's data base and further processed. A novel speed adaptation system for curves was invented, implemented and evaluated. This speed adaptation system is able to compute a vehicle speed suitable for the underlying driving style of the driver, road properties and weather conditions. Two different artificial neural networks and two mathematical models were used to compute the desired vehicle speed in curves. These methods were compared and evaluated.<br>Morgondagens fordon kommer att vara mer sofistikerade, intelligenta och säkra än dagens fordon. Framtiden lutar mot fullständigt autonoma fordon. Detta examensarbete tillhandahåller en datadriven lösning för ett hastighetsanpassningssystem som kan beräkna ett fordons hastighet i kurvor som är lämpligt för förarens körstil, vägens egenskaper och rådande väder. Ett hastighetsanpassningssystem för kurvor har som mål att beräkna en fordonshastighet för kurvor som kan användas i Advanced Driver Assistance Systems (ADAS) eller Autonomous Driving (AD) applikationer. Detta examensarbete utfördes på Volvo Car Corporation. Litteratur kring hastighetsanpassningssystem samt faktorer som påverkar ett fordons hastighet i kurvor studerades. Naturalistisk bilkörningsdata samlades genom att köra bil samt extraherades från Volvos databas och bearbetades. Ett nytt hastighetsanpassningssystem uppfanns, implementerades samt utvärderades. Hastighetsanpassningssystemet visade sig vara kapabelt till att beräkna en lämplig fordonshastighet för förarens körstil under rådande väderförhållanden och vägens egenskaper. Två olika artificiella neuronnätverk samt två matematiska modeller användes för att beräkna fordonets hastighet. Dessa metoder jämfördes och utvärderades.
27
Carrelli, David John. "Utilising Local Model Neural Network Jacobian Information in Neurocontrol." Thesis, 2006. http://hdl.handle.net/10539/1815.
Full textAbstract:
Student Number : 8315331 -
MSc dissertation -
School of Electrical and Information Engineering -
Faculty of Engineering and the Built Environment<br>In this dissertation an efficient algorithm to calculate the differential of the network output with respect to its inputs is derived for axis orthogonal Local Model (LMN) and Radial Basis Function (RBF) Networks. A new recursive Singular Value Decomposition (SVD) adaptation algorithm, which attempts to circumvent many of the problems found in existing recursive adaptation algorithms, is also derived. Code listings and simulations are presented to demonstrate how the algorithms may be used in on-line adaptive neurocontrol systems. Specifically, the control techniques known as series inverse neural control and instantaneous linearization are highlighted. The presented material illustrates how the approach enhances the flexibility of LMN networks making them suitable for use in both direct and indirect adaptive control methods. By incorporating this ability into LMN networks an important characteristic of Multi Layer Perceptron (MLP) networks is obtained whilst retaining the desirable properties of the RBF and LMN approach.
28
Carvalho, João Gabriel Marques. "Electricity consumption forecast model for the DEEC based on machine learning tools." Master's thesis, 2020. http://hdl.handle.net/10316/90148.
Full textAbstract:
Dissertação de Mestrado Integrado em Engenharia Electrotécnica e de Computadores apresentada à Faculdade de Ciências e Tecnologia<br>Nesta tese apresentaremos o trabalho sobre a criação de uma rede neuronal de aprendizagem automática, capaz de realizar previsões energéticas. Com o aumento do consumo energético, devem desenvolvidas ferramentas capazes de prever o consumo. Esta necessidade levou à pesquisa deste tema.Procura-se explicar a história da aprendizagem automática, o que é a aprendizagem automática e como é que esta funciona. Também se procura explicar os seus antecedentes matemáticos, a utilização de redes neuronais e que ferramentas foram atualmente desenvolvidas; de forma a criar soluções de aprendizagem automática. A aprendizagem automática consiste num programa informático, que após treino é capaz de desempenhar tarefas de forma similar à mente humana. A rede neuronal (ANN) é uma das mais importantes ferramentas de aprendizagem automática, através da qual se pode obter informação fundamental.Para prever o consumo de energia no Departamento de Engenharia Eletrotécnica e de Computadores (DEEC) da Universidade de Coimbra, uma rede neural foi treinada usando dados reais do consumo total das torres do DEEC.Phyton foi a linguagem utilizada e recorreu-se ao logaritmo de regressão de aprendizagem supervisionada. Com esta previsão, comparam-se os dados obtidos com os dados reais, o que permite a sua análise. Os dados usados no treino da rede neuronal vão de 2015/julho/10 a 2017/dezembro/31, num total de 906 dias. Por cada dia do ano existe um máximo de 3 valores, considerando-se assim uma amostra pequena.A comparação final entre os dados reais e os dados previstos foi somente realizada no mês de janeiro de 2018. A partir dos dados obtidos realizaram-se previsões, apesar de um certo nível de discrepância; justificada pela pequena quantidade de dados disponíveis. No futuro, deve-se aumentar os dados de treino de forma a obter um maior número de variáveis de entrada. O principal objetivo proposto nesta tese foi atingido com sucesso. Com toda a pesquisa apresentada, buscou-se criar informação que permitisse ser um marco na criação de melhores soluções. Este é um campo extraordinário que no futuro permitirá elevar os nossos conhecimentos a outros níveis.<br>In this thesis, the design of a machine learning neural network capable of making energy predictions is presented. With the increase in energy consumption, tools for the prediction of energy consumption are gaining great importance and their implementation is required. This concern is the main goal of the presented work.We strive to explain the history of machine learning, what machine learning is and how it works. It is also sought to explain the mathematical background and use of neural networks and what tools have been developed nowadays to create machine learning solutions. Machine learning is a computer program that can perform trained tasks in a similar way as the human mind. The neural network (ANN) is one of the most used and important machine learning solution through which pivotal data can be obtained. For predicting the energy consumption at the Department of Electrical and Computer Engineering (DEEC) of the University of Coimbra, a neural network was trained using real data from the overall consumption of the DEEC towers.Phyton was the language used and the supervised learning regression algorithm utilized. With this prediction, we finally compare our data with real data, so that we may analyze it. The data used in the training of the neural network goes from 2015/July/10 to 2017/December/31, a total of 906 days. For each day of the year, there is a maximum of 3 values, which is considered a small sample, but the only one available The final comparison between real and predicted data was only done for the month of January 2018. From the data achieved, predictions were made, but with a certain level of discrepancy, that is explained with the low amount of data available. In the future, one of the things that should be considered is to enlarge the training datasets, considering a larger amount of input variables. The main goal proposed for this thesis was successfully obtained. With all the presented research it was strived to create text that would allow being a steppingstone in the creation of better solutions. This is an extraordinary field that in the future will be able to elevate our knowledge to a completely different level.
29
Dubey, Abhishek. "Multimodal Deep Learning for Multi-Label Classification and Ranking Problems." Thesis, 2015. http://etd.iisc.ac.in/handle/2005/3681.
Full textAbstract:
In recent years, deep neural network models have shown to outperform many state of the art algorithms. The reason for this is, unsupervised pretraining with multi-layered deep neural networks have shown to learn better features, which further improves many supervised tasks. These models not only automate the feature extraction process but also provide with robust features for various machine learning tasks. But the unsupervised pretraining and feature extraction using multi-layered networks are restricted only to the input features and not to the output. The performance of many supervised learning algorithms (or models) depends on how well the output dependencies are handled by these algorithms [Dembczy´nski et al., 2012]. Adapting the standard neural networks to handle these output dependencies for any specific type of problem has been an active area of research [Zhang and Zhou, 2006, Ribeiro et al., 2012].
On the other hand, inference into multimodal data is considered as a difficult problem in machine learning and recently ‘deep multimodal neural networks’ have shown significant results [Ngiam et al., 2011, Srivastava and Salakhutdinov, 2012]. Several problems like classification with complete or missing modality data, generating the missing modality etc., are shown to perform very well with these models. In this work, we consider three nontrivial supervised learning tasks (i) multi-class classification (MCC),
(ii) multi-label classification (MLC) and (iii) label ranking (LR), mentioned in the order of increasing complexity of the output. While multi-class classification deals with predicting one class for every instance, multi-label classification deals with predicting more than one classes for every instance and label ranking deals with assigning a rank to each label for every instance. All the work in this field is associated around formulating new error functions that can force network to identify the output dependencies.
Aim of our work is to adapt neural network to implicitly handle the feature extraction (dependencies) for output in the network structure, removing the need of hand crafted error functions. We show that the multimodal deep architectures can be adapted for these type of problems (or data) by considering labels as one of the modalities. This also brings unsupervised pretraining to the output along with the input. We show that these models can not only outperform standard deep neural networks, but also outperform standard adaptations of neural networks for individual domains under various metrics over several data sets considered by us. We can observe that the performance of our models over other models improves even more as the complexity of the output/ problem increases.
30
Dubey, Abhishek. "Multimodal Deep Learning for Multi-Label Classification and Ranking Problems." Thesis, 2015. http://etd.iisc.ernet.in/2005/3681.
Full textAbstract:
In recent years, deep neural network models have shown to outperform many state of the art algorithms. The reason for this is, unsupervised pretraining with multi-layered deep neural networks have shown to learn better features, which further improves many supervised tasks. These models not only automate the feature extraction process but also provide with robust features for various machine learning tasks. But the unsupervised pretraining and feature extraction using multi-layered networks are restricted only to the input features and not to the output. The performance of many supervised learning algorithms (or models) depends on how well the output dependencies are handled by these algorithms [Dembczy´nski et al., 2012]. Adapting the standard neural networks to handle these output dependencies for any specific type of problem has been an active area of research [Zhang and Zhou, 2006, Ribeiro et al., 2012].
On the other hand, inference into multimodal data is considered as a difficult problem in machine learning and recently ‘deep multimodal neural networks’ have shown significant results [Ngiam et al., 2011, Srivastava and Salakhutdinov, 2012]. Several problems like classification with complete or missing modality data, generating the missing modality etc., are shown to perform very well with these models. In this work, we consider three nontrivial supervised learning tasks (i) multi-class classification (MCC),
(ii) multi-label classification (MLC) and (iii) label ranking (LR), mentioned in the order of increasing complexity of the output. While multi-class classification deals with predicting one class for every instance, multi-label classification deals with predicting more than one classes for every instance and label ranking deals with assigning a rank to each label for every instance. All the work in this field is associated around formulating new error functions that can force network to identify the output dependencies.
Aim of our work is to adapt neural network to implicitly handle the feature extraction (dependencies) for output in the network structure, removing the need of hand crafted error functions. We show that the multimodal deep architectures can be adapted for these type of problems (or data) by considering labels as one of the modalities. This also brings unsupervised pretraining to the output along with the input. We show that these models can not only outperform standard deep neural networks, but also outperform standard adaptations of neural networks for individual domains under various metrics over several data sets considered by us. We can observe that the performance of our models over other models improves even more as the complexity of the output/ problem increases.
31
Silvestre, Martinho de Matos. "Three-stage ensemble model : reinforce predictive capacity without compromising interpretability." Master's thesis, 2019. http://hdl.handle.net/10362/71588.
Full textAbstract:
Thesis proposal presented as partial requirement for obtaining the Master’s degree in Statistics and Information Management, with specialization in Risk Analysis and Management<br>Over the last decade, several banks have developed models to quantify credit risk. In addition to the monitoring of the credit portfolio, these models also help deciding the acceptance of new contracts, assess customers profitability and define pricing strategy. The objective of this paper is to improve the approach in credit risk modeling, namely in scoring models to predict default events. To this end, we propose the development of a three-stage ensemble model that combines the results interpretability of the Scorecard with the predictive power of machine learning algorithms. The results show that ROC index improves 0.5%-0.7% and Accuracy 0%-1% considering the Scorecard as baseline.
32
Liang, Ching-Hau, and 梁鈞皓. "Reconfigurable Neural Network Architectures Based on Multilayer Perceptron." Thesis, 2002. http://ndltd.ncl.edu.tw/handle/9jcr7b.
Full textAbstract:
碩士<br>逢甲大學<br>資訊工程所<br>90<br>Field Programmable Gate Array (FPGA) is an emerging hardware device recently. It make the products that produced by hardware designer enter the market quickly. And neural network is an important algorithm which recognition task needs. It is usually simulated by software, but the speed is not enough to satisfy the recognition task of real time. So, it is the best way to accelerate the neural network computing by implement with hardware. However, it lacks the resilience if implemented by ASIC. Therefore, the FPGA, which could adjust with different requirements,. Ex: number of neuron, number of input, and precision, becomes the best choice. But the capacity of current FPGA is not enough to fit a complex computing in it, therefore, we need to choice a tradeoff between spatial or temporal to perform the operation. This research is based on neural network multilayer with BEP learning to simulate the logic computing XOR, XNOR. We built a simulation environment to simulate the successful learning rate in different precision. And proposal several hardware architecture focus on spatial or temporal. Then build table of area or speed of several architecture. When we want to implement the network with constrains of spatial or temporal, we could find the appropriate architecture quickly and gain the best performance.
33
Weng, Zi-Jie, and 翁梓捷. "Using Multilayer Perceptron Neural Network to Assess the Critical Factors of Traffic Accidents." Thesis, 2018. http://ndltd.ncl.edu.tw/handle/4q4hqh.
Full textAbstract:
碩士<br>國立高雄第一科技大學<br>資訊管理系碩士班<br>106<br>In accordance with a report issued by the Ministry of Health and Welfare in 2017, it stated that death resulted from accident injury was the sixth fatal factor in 2016.Consequently, it is crucial to promote a better traffic prevention so as to avoid any possible traffic accidents. Nonetheless, due to the changing features of environment, drivers behaviors and others, the types of traffic accidents are diversified. This study is based on the data of traffic accidents in Taoyuan City in the government-opened data and the application of the data mining technology to construct an accident prediction model.
In the study, the data set of the traffic accidents in Taoyuan City was collected from the 2012 to 2017. The data was six classifiers were applied to predict and evaluate traffic accidents. The experimental results showed that: The used classifier were all be able to achieve good prediction results. In order to verify the stability of the classifier, cross-validation is also applied for model evaluation in the training process. Among them, the NN-MLP classifier presented outstandingly in the accuracy of the test dataset and the best performance evaluation of the model. In addition, to enhance experiments performance, boosting enhanced learning methods and discussion of the combination of factor traffic accidents were also applied. According to the above experiments, the results showed that using the Pearson Chi-square feature selection method to choose important traffic factor combinations and boosting methods can indeed help to improve the effectiveness of building traffic accidents model. The correction rate of NN-MLP model has reached 77% and 0.787 to AUC. While we constructed the model, the key factors we believe to cause the accidents are the degree of injury, the parts of the vehicle impact, the car models and the types of the accident, the main causes, the types of vehicle, and the time periods.
34
Gomes, Eduardo. "An agent-based approach to model farmers' land use cover change intentions." Doctoral thesis, 2019. http://hdl.handle.net/10451/48481.
Full textAbstract:
Land Use and Cover Change (LUCC) occurs as a consequence of both natural and human activities, causing impacts on biophysical and agricultural resources. In enlarged urban regions, the major changes are those that occur from agriculture to urban uses. Urban uses compete with rural ones due among others, to population growth and housing demand. This competition and the rapid nature of change can lead to fragmented and scattered land use development generating new challenges, for example, concerning food security, soil and biodiversity preservation, among others.
Landowners play a key role in LUCC. In peri-urban contexts, three interrelated key actors are pre-eminent in LUCC complex process: 1) investors or developers, who are waiting to take advantage of urban development to obtain the highest profit margin. They rely on population growth, housing demand and spatial planning strategies; 2) farmers, who are affected by urban development and intend to capitalise on their investment, or farmers who own property for amenity and lifestyle values; 3) and at a broader scale, land use planners/ decision-makers.
Farmers’ participation in the real estate market as buyers, sellers or developers and in the land renting market has major implications for LUCC because they have the capacity for financial investment and to control future agricultural land use.
Several studies have analysed farmer decision-making processes in peri-urban regions. These studies identified agricultural areas as the most vulnerable to changes, and where farmers are presented with the choice of maintaining their agricultural activities and maximising the production potential of their crops or selling their farmland to land investors. Also, some evaluate the behavioural response of peri-urban farmers to urban development, and income from agricultural production, agritourism, and off-farm employment. Uncertainty about future land profits is a major motivator for decisions to transform farmland into urban development. Thus, LUCC occurs when the value of expected urban development rents exceeds the value of agricultural ones. Some studies have considered two main approaches in analysing farmer decisions: how drivers influence farmer’s decisions; and how their decisions influence LUCC.
To analyse farmers’ decisions is to acknowledge the present and future trends and their potential spatial impacts. Simulation models, using cellular automata (CA), artificial neural networks (ANN) or agent-based systems (ABM) are commonly used.
This PhD research aims to propose a model to understand the agricultural land-use change in a peri-urban context. We seek to understand how human drivers (e.g., demographic, economic, planning) and biophysical drivers can affect farmer’s intentions regarding the future agricultural land and model those intentions. This study presents an exploratory analysis aimed at understanding the complex dynamics of LUCC based on farmers’ intentions when they are faced with four scenarios with the time horizon of 2025: the A0 scenario – based on current demographic, social and economic trends and investigating what happens if conditions are maintained (BAU); the A1 scenario – based on a regional food security; the A2 scenario – based on climate change; and the B0 scenario – based on farming under urban pressure, and investigating what happens if people start to move to rural areas. These scenarios were selected because of the early urbanisation of the study area, as a consequence of economic, social and demographic development; and because of the interest in preserving and maintaining agriculture as an essential resource. Also, Torres Vedras represents one of the leading suppliers of agricultural goods (mainly fresh fruits, vegetables, and wine) in Portugal.
To model LUCC a CA-Markov, an ANN-multilayer perceptron, and an ABM approach were applied. Our results suggest that significant LUCC will occur depending on farmers’ intentions in different scenarios. The highlights are: (1) the highest growth in permanently irrigated land in the A1 scenario; (2) the most significant drop in non-irrigated arable land, and the highest growth in the forest and semi-natural areas in the A2 scenario; and (3) the greatest urban growth was recognised in the B0 scenario. To verify if the fitting simulations performed well, statistical analysis to measure agreement and quantity-allocation disagreements and a participatory workshop with local stakeholders to validate the achieved results were applied. These outcomes could provide decision-makers with the capacity to observe different possible futures in ‘what if’ scenarios, allowing them to anticipate future uncertainties, and consequently allowing them the possibility to choose the more desirable future.
35
Chen, Ching-Chang, and 陳清泉. "Applying the method of multilayer perceptron neural network on high confused Mandarin vowel recognition." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/34273138390578821085.
Full textAbstract:
碩士<br>國立中興大學<br>統計學研究所<br>105<br>This thesis is mainly to apply the method of multilayer perceptron on two types of high confused Mandarin vowel recognition like <ㄣ、ㄥ>、<ㄛ、ㄨㄛ>、<ㄥ、ㄨㄥ> and so on.First, we would record the voices of twenty different speakers, and preprocess the voice signal like Digiting, Frame Blocking, Windowing and so on. Then MFCC are taken as the features of the voice signal, and the features of training data would be served as the input data to train the model. After training, input the test data and get the output of multilayer perceptron as the recognition rate of vowel. We use the vowel of Mandarin as data about 16000 on average each type which has 25 frames with 39 features each in total. Train the multilayer perceptron model with one or two hidden layer. The number of training voices is 7, 9, and 10 respectively. The results of experiment show that the error rate with 7 training voices is between 1.8 to 19.23 percent , the error rate with 9 training voices is between 1.44 and 18.73 percent, and the error rate with 10 training voices attain 0 percent. Finally, the results of the proposed method would compare with the results of KNN.
36
Xu, Teng Ren, and 許騰仁. "The research of applying the generalized perceptron neural network model to speech recognition." Thesis, 1994. http://ndltd.ncl.edu.tw/handle/72955544749627294616.
Full text
37
Kumar, Rajan. "A Neural Network Approach To Rotorcraft Parameter Estimation." Thesis, 2007. https://etd.iisc.ac.in/handle/2005/549.
Full textAbstract:
The present work focuses on the system identification method of aerodynamic parameter estimation which is used to calculate the stability and control derivatives required for aircraft flight mechanics. A new rotorcraft parameter estimation technique is proposed which uses a type of artificial neural network (ANN) called radial basis function network (RBFN). Rotorcraft parameter estimation using ANN is an unexplored research topic and the earlier works in this area have used the output error, equation error and filter error methods which are conventional parameter estimation methods. However, the conventional methods require an accurate non-linear rotorcraft simulation model which is not required by the ANN based method. The application of RBFN overcomes the drawbacks of multilayer perceptron (MLP) based delta method of parameter estimation and gives satisfactory results at either end of the ordered set of estimates. This makes the RBFN based delta method for parameter estimation suitable for rotorcraft studies, as both transition and high speed flight regime characteristics can be studied. The RBFN based delta method for parameter estimation is used for computation of aerodynamic parameters from both simulated and real time flight data. The simulated data is generated from an 8-DoF non-linear simulation model based on the Level-1 criteria of rotorcraft simulation modeling. The generated simulated data is used for computation of the quasi-steady and the time-variant stability and control parameters for different flight conditions using the RBFN based delta method. The performance of RBFN based delta method is also analyzed in the presence of state and measurement noise as well as outliers. The established methodology is then applied to compute parameters directly from real time flight test data for a BO 105 S123 helicopter obtained from DLR (German Aerospace Center). The parameters identified using the RBFN based delta method are compared with the identified values for the BO 105 helicopter from published literature which have used conventional parameter estimation techniques for parameter estimation using a 6-DoF and a 9-DoF rotorcraft simulation model. Finally, the estimated parameters are verified from the flight data generated by a frequency sweep pilot control input for assessing the predictive capability of the RBFN based delta method. Since the approach directly computes the parameters from flight data, it can be used for a reliable description of the higher frequency range, which is needed for high bandwidth flight control and in-flight simulation.
38
Kumar, Rajan. "A Neural Network Approach To Rotorcraft Parameter Estimation." Thesis, 2007. http://hdl.handle.net/2005/549.
Full textAbstract:
The present work focuses on the system identification method of aerodynamic parameter estimation which is used to calculate the stability and control derivatives required for aircraft flight mechanics. A new rotorcraft parameter estimation technique is proposed which uses a type of artificial neural network (ANN) called radial basis function network (RBFN). Rotorcraft parameter estimation using ANN is an unexplored research topic and the earlier works in this area have used the output error, equation error and filter error methods which are conventional parameter estimation methods. However, the conventional methods require an accurate non-linear rotorcraft simulation model which is not required by the ANN based method. The application of RBFN overcomes the drawbacks of multilayer perceptron (MLP) based delta method of parameter estimation and gives satisfactory results at either end of the ordered set of estimates. This makes the RBFN based delta method for parameter estimation suitable for rotorcraft studies, as both transition and high speed flight regime characteristics can be studied. The RBFN based delta method for parameter estimation is used for computation of aerodynamic parameters from both simulated and real time flight data. The simulated data is generated from an 8-DoF non-linear simulation model based on the Level-1 criteria of rotorcraft simulation modeling. The generated simulated data is used for computation of the quasi-steady and the time-variant stability and control parameters for different flight conditions using the RBFN based delta method. The performance of RBFN based delta method is also analyzed in the presence of state and measurement noise as well as outliers. The established methodology is then applied to compute parameters directly from real time flight test data for a BO 105 S123 helicopter obtained from DLR (German Aerospace Center). The parameters identified using the RBFN based delta method are compared with the identified values for the BO 105 helicopter from published literature which have used conventional parameter estimation techniques for parameter estimation using a 6-DoF and a 9-DoF rotorcraft simulation model. Finally, the estimated parameters are verified from the flight data generated by a frequency sweep pilot control input for assessing the predictive capability of the RBFN based delta method. Since the approach directly computes the parameters from flight data, it can be used for a reliable description of the higher frequency range, which is needed for high bandwidth flight control and in-flight simulation.
39
Oba, Pius Nwachukwu. "Quality analysis modelling for development of a process controller in resistance spot welding using neural networks techniques." Thesis, 2006. http://hdl.handle.net/10539/1676.
Full textAbstract:
Student Number : 9811923K -
PhD thesis -
School of Mechanical Engineering -
Faculty of Engineering and the Built Environment<br>Methods are presented for obtaining models used for predicting welded sample resistance and effective weld current (RMS) for desired weld diameter (weld quality) in the resistance spot welding process. These models were used to design predictive controllers for the welding process. A suitable process model forms an important step in the development and design of process controllers for achieving good weld quality with good reproducibility.
Effective current, dynamic resistance and applied electrode force are identified as important input parameters necessary to predict the output weld diameter. These input parameters are used for the process model and design of a predictive controller.
A three parameter empirical model with dependent and independent variables was used for curve fitting the nonlinear halfwave dynamic resistance. The estimates of the parameters were used to develop charts for determining overall resistance of samples for any desired weld diameter. Estimating resistance for samples welded in the machines from which dataset obtained were used to plot the chart yielded accurate results. However using these charts to estimate sample resistance for new and unknown machines yielded high estimation error. To improve the prediction accuracy the same set of data generated from the model were used to train four different neural network types. These were the Generalised Feed Forward (GFF) neural network, Multilayer Perceptron (MLP) network, Radial Basis Function (RBF) and Recurrent neural network (RNN).
Of the four network types trained, the MLP had the least mean square error for training and cross validation of 0.00037 and 0.00039 respectively with linear correlation coefficient in testing of 0.999 and maximum estimation error range from 0.1% to 3%. A prediction accuracy of about 97% to 99.9%. This model was selected for the design and implementation of the controller for predicting overall sample resistance. Using this predicted overall sample resistance, and applied electrode force, a second model was developed for predicting required effective weld current for any desired weld diameter. The prediction accuracy of this model was in the range of 94% to 99%.
The neural network predictive controller was designed using the MLP neural network models. The controller outputs effective current for any desired weld diameter and is observed to track the desired output accurately with same prediction accuracy of the model used which was about 94% to 99%. The controller works by utilizing the neural network output embedded in Microsoft Excel as a digital link library and is able to generate outputs for given inputs on activating the process by the push of a command button.
40
Teixeira, Rafael Gonçalves. "Energy disaggregation using Machine Learning." Master's thesis, 2021. http://hdl.handle.net/10773/33657.
Full textAbstract:
Nowadays, we are surrounded by electric appliances. Either at home by the washing
machine, kettle, or oven, or work by the computer, cellphone, or printer. Such
devices help us daily, but their popularization increased the energy consumption
to concerning values. In an attempt to reduce energy consumption, governments
started enforcing policies regarding energy education to teach homeowners how to
reduce energy wastage on the demand side. One of those policies was the deployment
of smart meters, which allow the consumer to know how much energy is
being consumed at any given time through a display on the household energy meter.
Even though this measure was well received, the studies show that the best results
in energy conservation are obtained through real-time appliance level feedback. To
get such feedback, one can either measure every outlet in a household, which is
unviable for a broad deployment solution, or disaggregate the energy recorded by
the smart meter. NILM or Non-Intrusive Load Monitoring is the name we give to
the second option where we use the aggregated readings of a household to find the
energy consumed by each appliance. There were many proposals to solve NILM
ranging from HMMs to GSP, where deep learning models showed remarkable results,
obtaining state-of-the-art results. With the intent to create a complete NILM
solution, Withus partnered with the University of Aveiro and proposed this dissertation.
The initial objective was to develop a machine learning model to solve
NILM. Still, during the background analysis, we found the need to create a new
dataset which led to the expansion of the initial proposal to include the dataset
preprocessing and conversion. Regarding NILM, we proposed three new deep learning
models: a convolutional neural network with residual blocks, a recurrent neural
network, and a multilayer perceptron that uses discrete wavelet transforms as features.
These models went through multiple iterations, being evaluated first in the
simpler ON/OFF classification task and later modified and evaluated for the disaggregation
task. We compared our models to the state-of-the-art ones proposed in
NILMTK, where they presented better results than the real-time alternative, dAE,
reducing the NRMSE on average by 49%. We also got close to the best option that
classified with a 30 min delay, Seq2Point, increasing the error on average by 17%.
Besides that, we also analyze the best models from the previous comparison on the
benefit of transfer learning between datasets, where the results show a marginal
performance improvement when using transfer learning. This document presents
the solution outline definition, the multiple options considered for dataset processing
and the best solution, the models’ evolution and results, and the comparison
with the state-of-the-art models regarding generalization to different houses and
under transfer learning.<br>Hoje em dia estamos rodeados de dispositivos elétricos. Quer seja em casa, pela
máquina de lavar, o microondas ou o forno ou no emprego pelo computador, o
telemóvel ou a impressora. Estes dispositivos ajudam-nos diariamente, mas com
a sua popularização o consumo energético atingiu valores preocupantes. Numa
tentativa de reduzir o consumo energético, os governos começaram a introduzir
políticas de educação energética para ensinar os consumidores a reduzir o desperdício
energético. Uma das medidas foi a implementação generalizada de smart
meters, que permitem ao consumidor saber quanta energia está a ser consumida
a qualquer altura através de um ecrã no contador da casa. Mesmo sendo bem
recebida, esta medida não é suficiente uma vez que os estudos indicam que os
melhores resultados são obtidos através de feedback ao nível do dispositivo em
tempo real. Para obtermos este feedback existem duas formas, podemos medir
cada tomada numa dada casa, o que é inviável para uma implementação em larga
escala, ou desagregar a energia registrada pelo smart meter que já está presente
na casa. NILM ou Non-Intrusive Load Monitoring é o nome dado à segunda opção
onde a energia agregada da casa é usada para descobrirmos a energia consumida
por cada dispositivo elétrico. Para resolver este problema foram propostas várias
alternativas, desde HMMs a GSP, onde os modelos de deep learning obtiveram
resultados notáveis sendo agora o estado da arte. Com o objetivo de produzir um
sistema NILM completo, a Withus juntou-se à Universidade de Aveiro e juntos propuseram
esta dissertação. O objetivo inicial era o desenvolvimento de um modelo
de machine learning para desagregar consumos elétricos. Contudo, durante análise
do estado da arte, deparamo-nos com a necessidade de criar um novo dataset, o
que levou à extensão da proposta inicial para incluir também o pré-processamento
e conversão do dataset. Para desagregação de consumos elétricos propusemos três
modelos: uma rede neuronal convolucional com blocos residuais, uma rede neuronal
recorrente e um multilayer perceptron que usa discrete wavelet transforms
como features. Estes modelos passaram por diversas iterações, sendo avaliados
primeiro na tarefa de classificação ON/OFF e depois modificados e avaliados para
desagregação. Os modelos foram ainda comparados com os do estado da arte
presentes no NILMTK, onde apresentaram melhores resultados que a alternativa
real-time, dAE, diminuindo o NRMSE em média 49% ficando próximos da melhor
alternativa que classifica com atraso, Seq2Point, apresentando um erro pior, em
média, de 17%. Para além disso, também analisamos os melhores modelos da
experiência anterior no benefício de usar transfer learning entre datasets, onde os
resultados mostram uma melhoria marginal quando usamos transfer learning. Este
documento apresenta a definição do esboço da solução, as múltiplas opções consideradas
para processamento de dataset e qual a melhor, a evolução dos modelos,
os seus resultados e a comparação com os modelos do estado da arte na capacidade
de generalização entre diferentes casas e de transfer learning entre datasets.<br>Mestrado em Engenharia Informática
41
PROSPERI, MAURIZIO. "Analysis of the EU MacSharry Reform in Italian Family Farms: A Multilayer Feed-Forward Neural Network model for evaluating economic and environmental impacts of the direct payments system." Doctoral thesis, 2005. http://hdl.handle.net/1234/13260.
Full textAbstract:
Nella UE è stata per la prima volta introdotta nel 1992 una radicale riforma della PAC riguardanti diverse coltivazioni, consistente in una riduzione del prezzo istituzionale ed in un sistema di sostegno diretto al reddito. In questo articolo, viene verificato l’impatto provocato dalla riforma della PAC in aziende agricole operanti in diverse condizioni geografiche della regione Marche, in Italia, al fine di quantificare la variazione di reddito agricolo provocato dalla riforma. In merito alla metodologia, è stato applicato un modello di tipo Multilayer Feedforwad Neural Network (MFNN) che, rispetto ai modelli statistici più comuni, risulta più effettivo qualora venga applicato sui dati direttamente rilevati dalle aziende agricole. In seguito all’applicazione di questo modello, si sono potuti utilizzare i dati rilevati dall’imponente rete di rilevazione di dati aziendali (FADN) presente nell’UE, consentendo di condurre una analisi comparativa tra diverse aree geografiche. Dall’applicazione del modello MFNN, è stato possibile calcolare l’effetto disaggregato della riduzione del prezzo istituzionale e dell’introduzione del sistema di pagamenti diretti. Dai risultati è emerso che nelle aree fertili si è verificato una compensazione al reddito eccessiva. La ragione di ciò è dovuta al metodo di determinazione delle compensazioni finora adottato nei confronti di ogni azienda, che causa l’insorgere di discriminazioni tra aree agricole affette da diverse condizioni operative. Infatti, il metodo di calcolo adottato dall’UE considera la resa media delle colture per ettaro ma, al fine di migliorare l’equità della distribuzione, si dovrebbero provare metodi di calcolo alternativi al fine di ricercarne uno più appropriato per la determinazione dei pagamenti diretti.
42
Sousa, Martim Afonso Gouveia. "Previsão de vendas de peças de substituição para otimização da gestão de stock." Master's thesis, 2021. http://hdl.handle.net/10773/32749.
Full textAbstract:
Este projeto tem como intuito prever o número de vendas de peças de substituição,
associadas a equipamentos de termotecnologia, das próximas três semanas
para a empresa Bosch Termotecnologia-Aveiro. Desta forma, a Bosch pode otimizar
o planeamento de produção e stocks garantindo a satisfação do cliente
final.
A nossa abordagem ao problema consiste no estudo das séries temporais de vendas
destas peças de substituição, construindo modelos matemáticos capazes de
captar os padrões existentes nos dados. Os modelos utilizados cobrem um largo
espetro de metodologias, por forma a acomodar diferentes tipos de séries temporais.
Concretamente, utilizámos metodologias clássicas como o Autoregressive
integrated moving average (ARIMA) e Holt-Winters, mas também métodos
de aprendizagem supervisionada como o Support vector regression (SVR) e Extreme
gradient boosting (XGBoost) e ainda modelos de redes neuronais, nomeadamente
Multilayer perceptron (MLP), Convolutional neural network (CNN) e
Long short-term memory (LSTM).<br>This project aims to forecast the number of sales of spare parts, linked to
thermotechnology equipment, in the next three weeks for the company Bosch
Thermotechnology-Aveiro. In this way, Bosch can optimize production planning,
reducing production and storage costs.
Our approach to the problem consists of studying the sales time series of these
spare parts, in order to build mathematical models capable of capturing the patterns
underlying in the data. The models used cover a wide spectrum of methodologies,
with the purpose of fitting different types of time series. Specifically,
we used classical methodologies such as ARIMA and Holt-Winters, but also supervised
learning methods such as SVR and XGBoost and also deep learning
approaches, namely MLP, CNN and LSTM.<br>Mestrado em Matemática e Aplicações
43
Василенко, Дмитро Олексійович. "Конструктивний синтез планарних антен природними алгоритмами оптимізації". Doctoral thesis, 2010. https://ela.kpi.ua/handle/123456789/641.
Full text