Academic literature on the topic 'Neural networks (NNs)'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the lists of relevant articles, books, theses, conference reports, and other scholarly sources on the topic 'Neural networks (NNs).'
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.
Journal articles on the topic "Neural networks (NNs)"
1
Thakur, Amey. "Fundamentals of Neural Networks." International Journal for Research in Applied Science and Engineering Technology 9, no. VIII (August 15, 2021): 407–26. http://dx.doi.org/10.22214/ijraset.2021.37362.
Full textAbstract:
The purpose of this study is to familiarise the reader with the foundations of neural networks. Artificial Neural Networks (ANNs) are algorithm-based systems that are modelled after Biological Neural Networks (BNNs). Neural networks are an effort to use the human brain's information processing skills to address challenging real-world AI issues. The evolution of neural networks and their significance are briefly explored. ANNs and BNNs are contrasted, and their qualities, benefits, and disadvantages are discussed. The drawbacks of the perceptron model and their improvement by the sigmoid neuron and ReLU neuron are briefly discussed. In addition, we give a bird's-eye view of the different Neural Network models. We study neural networks (NNs) and highlight the different learning approaches and algorithms used in Machine Learning and Deep Learning. We also discuss different types of NNs and their applications. A brief introduction to Neuro-Fuzzy and its applications with a comprehensive review of NN technological advances is provided.
2
Jwo, Dah-Jing, and Chien-Cheng Lai. "Neural Network-Based Geometry Classification for Navigation Satellite Selection." Journal of Navigation 56, no. 2 (May 2003): 291–304. http://dx.doi.org/10.1017/s0373463303002200.
Full textAbstract:
The neural networks (NN)-based geometry classification for good or acceptable navigation satellite subset selection is presented. The approach is based on classifying the values of satellite Geometry Dilution of Precision (GDOP) utilizing the classification-type NNs. Unlike some of the NNs that approximate the function, such as the back-propagation neural network (BPNN), the NNs here are employed as classifiers. Although BPNN can also be employed as a classifier, it takes a long training time. Two other methods that feature a fast learning speed will be implemented, including Optimal Interpolative (OI) Net and Probabilistic Neural Network (PNN). Simulation results from these three neural networks are presented. The classification performance and computational expense of neural network-based GDOP classification are explored.
3
Guidotti, Dario. "Verification and Repair of Neural Networks." Proceedings of the AAAI Conference on Artificial Intelligence 35, no. 18 (May 18, 2021): 15714–15. http://dx.doi.org/10.1609/aaai.v35i18.17854.
Full textAbstract:
Neural Networks (NNs) are popular machine learning models which have found successful application in many different domains across computer science. However, it is hard to provide any formal guarantee on the behaviour of neural networks and therefore their reliability is still in doubt, especially concerning their deployment in safety and security-critical applications. Verification emerged as a promising solution to address some of these problems. In the following, I will present some of my recent efforts in verifying NNs.
4
IKEDA, TAKASHI, and MASAFUMI HAGIWARA. "CONTENT-BASED IMAGE RETRIEVAL SYSTEM USING NEURAL NETWORKS." International Journal of Neural Systems 10, no. 05 (October 2000): 417–24. http://dx.doi.org/10.1142/s0129065700000326.
Full textAbstract:
An effective image retrieval system is developed based on the use of neural networks (NNs). It takes advantages of association ability of multilayer NNs as matching engines which calculate similarities between a user's drawn sketch and the stored images. The NNs memorize pixel information of every size-reduced image (thumbnail) in the learning phase. In the retrieval phase, pixel information of a user's drawn rough sketch is inputted to the learned NNs and they estimate the candidates. Thus the system can retrieve candidates quickly and correctly by utilizing the parallelism and association ability of NNs. In addition, the system has learning capability: it can automatically extract features of a user's drawn sketch during the retrieval phase and can store them as additional information to improve the performance. The software for querying, including efficient graphical user interfaces, has been implemented and tested. The effectiveness of the proposed system has been investigated through various experimental tests.
5
Ghorbani, Behrooz, Song Mei, Theodor Misiakiewicz, and Andrea Montanari. "When do neural networks outperform kernel methods?*." Journal of Statistical Mechanics: Theory and Experiment 2021, no. 12 (December 1, 2021): 124009. http://dx.doi.org/10.1088/1742-5468/ac3a81.
Full textAbstract:
Abstract For a certain scaling of the initialization of stochastic gradient descent (SGD), wide neural networks (NN) have been shown to be well approximated by reproducing kernel Hilbert space (RKHS) methods. Recent empirical work showed that, for some classification tasks, RKHS methods can replace NNs without a large loss in performance. On the other hand, two-layers NNs are known to encode richer smoothness classes than RKHS and we know of special examples for which SGD-trained NN provably outperform RKHS. This is true even in the wide network limit, for a different scaling of the initialization. How can we reconcile the above claims? For which tasks do NNs outperform RKHS? If covariates are nearly isotropic, RKHS methods suffer from the curse of dimensionality, while NNs can overcome it by learning the best low-dimensional representation. Here we show that this curse of dimensionality becomes milder if the covariates display the same low-dimensional structure as the target function, and we precisely characterize this tradeoff. Building on these results, we present the spiked covariates model that can capture in a unified framework both behaviors observed in earlier work. We hypothesize that such a latent low-dimensional structure is present in image classification. We test numerically this hypothesis by showing that specific perturbations of the training distribution degrade the performances of RKHS methods much more significantly than NNs.
6
Đerek, Jurica, Marjan Sikora, Luka Kraljević, and Mladen Russo. "Using Neural Networks for Bicycle Route Planning." Applied Sciences 11, no. 21 (October 27, 2021): 10065. http://dx.doi.org/10.3390/app112110065.
Full textAbstract:
This paper presents the usage of artificial neural networks (NNs) in bicycle route planning. This research aimed to check the possibility of NNs to transfer human expertise in bicycle route design by training the NN on an already established set of bicycle routes and then using the trained NN to design the routes on the novel area. We created two NNs capable of choosing the best route among the given road network by training them on two different areas. The bicycle routes produced by NNs were the same at best and had 75% overlap at the worst compared to those produced by human experts. Furthermore, the mean square error for all of our NN models varied from 0.015 and 0.081. We compared this new approach to the traditional multicriteria GIS (geographic information system) analysis (MA) that requires the human expert to define the bicycle route selection criteria. The benefit of using NN over the MA was that the NN directly transfers the human expertise to a model. In contrast, the MA needs the expert to select multiple criteria and adjust their weights carefully.
7
Hu, Yibo, Yuzhe Ou, Xujiang Zhao, Jin-Hee Cho, and Feng Chen. "Multidimensional Uncertainty-Aware Evidential Neural Networks." Proceedings of the AAAI Conference on Artificial Intelligence 35, no. 9 (May 18, 2021): 7815–22. http://dx.doi.org/10.1609/aaai.v35i9.16954.
Full textAbstract:
Traditional deep neural networks (NNs) have significantly contributed to the state-of-the-art performance in the task of classification under various application domains. However, NNs have not considered inherent uncertainty in data associated with the class probabilities where misclassification under uncertainty may easily introduce high risk in decision making in real-world contexts (e.g., misclassification of objects in roads leads to serious accidents). Unlike Bayesian NN that indirectly infer uncertainty through weight uncertainties, evidential NNs (ENNs) have been recently proposed to explicitly model the uncertainty of class probabilities and use them for classification tasks. An ENN offers the formulation of the predictions of NNs as subjective opinions and learns the function by collecting an amount of evidence that can form the subjective opinions by a deterministic NN from data. However, the ENN is trained as a black box without explicitly considering inherent uncertainty in data with their different root causes, such as vacuity (i.e., uncertainty due to a lack of evidence) or dissonance (i.e., uncertainty due to conflicting evidence). By considering the multidimensional uncertainty, we proposed a novel uncertainty-aware evidential NN called WGAN-ENN (WENN) for solving an out-of-distribution (OOD) detection problem. We took a hybrid approach that combines Wasserstein Generative Adversarial Network (WGAN) with ENNs to jointly train a model with prior knowledge of a certain class, which has high vacuity for OOD samples. Via extensive empirical experiments based on both synthetic and real-world datasets, we demonstrated that the estimation of uncertainty by WENN can significantly help distinguish OOD samples from boundary samples. WENN outperformed in OOD detection when compared with other competitive counterparts.
8
Krogh, Anders, and Søren Kamaric Riis. "Hidden Neural Networks." Neural Computation 11, no. 2 (February 1, 1999): 541–63. http://dx.doi.org/10.1162/089976699300016764.
Full textAbstract:
A general framework for hybrids of hidden Markov models (HMMs) and neural networks (NNs) called hidden neural networks (HNNs) is described. The article begins by reviewing standard HMMs and estimation by conditional maximum likelihood, which is used by the HNN. In the HNN, the usual HMM probability parameters are replaced by the outputs of state-specific neural networks. As opposed to many other hybrids, the HNN is normalized globally and therefore has a valid probabilistic interpretation. All parameters in the HNN are estimated simultaneously according to the discriminative conditional maximum likelihood criterion. The HNN can be viewed as an undirected probabilistic independence network (a graphical model), where the neural networks provide a compact representation of the clique functions. An evaluation of the HNN on the task of recognizing broad phoneme classes in the TIMIT database shows clear performance gains compared to standard HMMs tested on the same task.
9
DeVore, Ronald, Boris Hanin, and Guergana Petrova. "Neural network approximation." Acta Numerica 30 (May 2021): 327–444. http://dx.doi.org/10.1017/s0962492921000052.
Full textAbstract:
Neural networks (NNs) are the method of choice for building learning algorithms. They are now being investigated for other numerical tasks such as solving high-dimensional partial differential equations. Their popularity stems from their empirical success on several challenging learning problems (computer chess/Go, autonomous navigation, face recognition). However, most scholars agree that a convincing theoretical explanation for this success is still lacking. Since these applications revolve around approximating an unknown function from data observations, part of the answer must involve the ability of NNs to produce accurate approximations.This article surveys the known approximation properties of the outputs of NNs with the aim of uncovering the properties that are not present in the more traditional methods of approximation used in numerical analysis, such as approximations using polynomials, wavelets, rational functions and splines. Comparisons are made with traditional approximation methods from the viewpoint of rate distortion, i.e. error versus the number of parameters used to create the approximant. Another major component in the analysis of numerical approximation is the computational time needed to construct the approximation, and this in turn is intimately connected with the stability of the approximation algorithm. So the stability of numerical approximation using NNs is a large part of the analysis put forward.The survey, for the most part, is concerned with NNs using the popular ReLU activation function. In this case the outputs of the NNs are piecewise linear functions on rather complicated partitions of the domain of f into cells that are convex polytopes. When the architecture of the NN is fixed and the parameters are allowed to vary, the set of output functions of the NN is a parametrized nonlinear manifold. It is shown that this manifold has certain space-filling properties leading to an increased ability to approximate (better rate distortion) but at the expense of numerical stability. The space filling creates the challenge to the numerical method of finding best or good parameter choices when trying to approximate.
10
Wongsathan, Rati, and Pasit Pothong. "Heart Disease Classification Using Artificial Neural Networks." Applied Mechanics and Materials 781 (August 2015): 624–27. http://dx.doi.org/10.4028/www.scientific.net/amm.781.624.
Full textAbstract:
Neural Networks (NNs) has emerged as an importance tool for classification in the field of decision making. The main objective of this work is to design the structure and select the optimized parameter in the neural networks to implement the heart disease classifier. Three types of neural networks, i.e. Multi-layered Perceptron Neural Network (MLP-NN), Radial Basis Function Neural Networks (RBF-NN), and Generalized Regression Neural Network (GR-NN) have been used to test the performance of heart disease classification. The classification accuracy obtained by RBFNN gave a very high performance than MLP-NN and GR-NN respectively. The performance of accuracy is very promising compared with the previously reported another type of neural networks.
Dissertations / Theses on the topic "Neural networks (NNs)"
1
Fischer, Manfred M. "Neural networks. A class of flexible non-linear models for regression and classification." Elgar, 2015. http://epub.wu.ac.at/4763/1/NN%2DHandbook%2Dchapter_Fischer_20120809.pdf.
Full text
2
Zirpe, Milind A. "RAIN and NCS 5 benchmarks." abstract and full text PDF (free order & download UNR users only), 2007. http://0-gateway.proquest.com.innopac.library.unr.edu/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqdiss&rft_dat=xri:pqdiss:1447612.
Full text
3
Mattos, CÃsar Lincoln Cavalcante. "ComitÃs de Classificadores Baseados nas Redes SOM e Fuzzy ART com Sintonia de ParÃmetros e SeleÃÃo de Atributos via MetaheurÃsticas EvolucionÃrias." Universidade Federal do CearÃ, 2011. http://www.teses.ufc.br/tde_busca/arquivo.php?codArquivo=7034.
Full textAbstract:
CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel SuperiorO paradigma de classificaÃÃo baseada em comitÃs tem recebido considerÃvel atenÃÃo na literatura cientÃfica em anos recentes. Neste contexto, redes neurais supervisionadas tÃm sido a escolha mais comum para compor os classificadores base dos comitÃs. Esta dissertaÃÃo tem a intenÃÃo de projetar e avaliar comitÃs de classificadores obtidos atravÃs de modificaÃÃes impostas a algoritmos de aprendizado nÃo-supervisionado, tais como as redes Fuzzy ART e SOM, dando origem, respectivamente, Ãs arquiteturas ARTIE (ART in Ensembles) e MUSCLE (Multiple SOM Classifiers in Ensembles). A sintonia dos parÃmetros e a seleÃÃo dos atributos das redes neurais que compÃem as arquiteturas ARTIE e MUSCLE foram tratados por otimizaÃÃo metaheurÃstica, a partir da proposiÃÃo do algoritmo I-HPSO (Improved Hybrid Particles Swarm Optimization). As arquiteturas ARTIE e MUSCLE foram avaliadas e comparadas com comitÃs baseados nas redes Fuzzy ARTMAP, LVQ e ELM em 12 conjuntos de dados reais. Os resultados obtidos indicam que as arquiteturas propostas apresentam desempenhos superiores aos dos comitÃs baseados em redes neurais supervisionadas.
4
Bezerra, Alberto Guilherme de Oliveira. "Modelos de previsão de tarifa de água, aplicados a autarquias municipais e empresas privadas, nas regiões Sul e Sudeste do Brasil /." Ilha Solteira, 2019. http://hdl.handle.net/11449/183655.
Full textAbstract:
Orientador: Marcelo LibânioResumo: O objetivo do presente trabalho é avaliar modelos de previsão de tarifa de água, aplicados a autarquias municipais e empresas privadas, nas regiões Sul e Sudeste do Brasil. Utilizando a metodologia de cálculo e posterior comparação dos erros obtidos para as previsões, verificando também a aplicabilidade das tarifas previstas para cada sistema de abastecimento. Utilizou-se dois modelos de previsão, o primeiro, fundamentado em técnicas de regressão linear múltipla e o segundo, baseado na aplicação de redes neurais artificiais. Avaliando, dessa forma, a capacidade de os dois modelos em questão preverem os valores tarifários a serem cobrados pelos prestadores de serviços de abastecimento de água e coleta de esgoto, a partir da análise das tarifas anteriormente praticadas. Os dados subsidiários para elaboração dos modelos foram obtidos por meio do sistema nacional de informações sobre saneamento (SNIS). Confirmada a consistência do banco de dados primário, procedeu-se com processamento destes dados, e definição das variáveis mais intervenientes para a definição da tarifa por meio da técnica de análise de correlação. Propôs-se a classificação dos sistemas de acordo com a classe jurídica do prestador de serviço, os cenários financeiros (superávit ou déficit) destes prestadores e o porte populacional dos municípios atendidos. Os resultados obtidos indicaram que os processos de previsão, em ambos os modelos utilizados, foram capazes de prever com elevada acurácia as tarifas, e garanti... (Resumo completo, clicar acesso eletrônico abaixo)
Abstract: The objective of the present work was evaluating forecasting models for water tariff applied to municipal and private companies in the South and Southeast regions of Brazil. Using the calculation methodology and subsequent comparison of the errors obtained for the forecasts, also verifying the applicability of the forecast tariffs for each supply system. Two prediction models are used, the first based on multiple linear regression techniques and the second based on the application of artificial neural networks. Evaluating, in this way, the ability of the two models in question to predict the tariff values to be charged by the water supply and wastewater collection service providers, based on the analysis of the tariffs previously practiced. The subsidiary data for the elaboration of the models were obtained through the national sanitation information system (SNIS). Confirming the consistency of the primary database, we proceeded with processing of these data and definition of the most intervening variables for the definition of the tariff through the correlation analysis technique. The classification of the systems according to the legal class of the service provider, the financial scenarios (surplus or deficit) of these providers and the population size of the municipalities served were proposed. The obtained results indicated that the forecasting processes, in both models used, were able to predict with high accuracy the tariffs, and guaranteed the maintenance of the surplu... (Complete abstract click electronic access below)
Mestre
5
Tavares, Guilherme Farias. "Modelagem matemática e sistemas inteligentes para predição do comportamento alimentar de suínos nas fases de crescimento e terminação." Universidade de São Paulo, 2017. http://www.teses.usp.br/teses/disponiveis/11/11152/tde-28072017-082242/.
Full textAbstract:
A suinocultura é uma atividade de grande importância em termos mundiais e de Brasil. Entretanto, por serem animais homeotérmicos, algumas alterações no ambiente térmico de alojamento podem alterar suas respostas fisiológicas e comportamentais para manutenção da temperatura interna. Portanto, o objetivo dessa pesquisa foi avaliar o comportamento alimentar de suínos, mediante a influência do ambiente térmico, nas fases de crescimento e terminação para diferentes linhagens comerciais e sexo. Além disso, buscou-se o desenvolvimento de modelos matemáticos e sistemas inteligentes para predição do tempo em alimentação (TM, min dia-1) dos suínos. Os dados foram coletados em uma granja experimental de suínos, localizada na cidade de Clay Center, Nebraska, Estados Unidos. O período experimental contemplou duas estações durante o ano 2015/2016 (verão e inverno), totalizando 63 dias (9 semanas) de informações coletadas para cada estação. Os animais alojados foram de três linhagens comerciais distintas: Landrace, Duroc e Yorkshire. Cada baia apresentava composição mista, sendo alojados 40 animais de diferentes linhagens comerciais e sexo. No total, foram confinados 240 animais, sendo 80 animais para cada linhagem comercial entre machos castrados e fêmeas. Foram registrados dados de temperatura do ar (Tar, °C), temperatura do ponto de orvalho (Tpo, °C) e umidade relativa do ar (UR, %) a cada 5 minutos no interior da instalação. Para TM, os dados foram coletados e registrados a cada 20 segundos por meio de um sistema de coleta de dados por rádio frequência. O conforto térmico foi analisado a partir do Índice de Temperatura e Umidade (ITU) e a Entalpia Específica (H, kJ kg-1 de ar seco). Para avaliar a relação entre o ambiente térmico e TM, foi utilizada estatística multivariada por meio de análise de componentes principais (ACP) e agrupamento para obtenção de padrões e seleção de variáveis para entrada nos modelos. O modelo fuzzy e as redes neurais artificias foram desenvolvidos em ambiente MATLAB® R2015a por meio dos toolboxes Fuzzy e Neural Network, com o objetivo de predizer TM, tendo como variáveis de entrada: linhagem comercial, sexo, idade e ITU. De uma maneira geral, as médias de Tar estiveram dentro da zona de termoneutralidade (ZCT) em todo período experimental, sendo que apenas a UR apresentou valores abaixo da UR crítica inferior. Para o ITU, apenas no verão foram encontrados valores acima da ZCT, entretanto, esses valores estiveram abaixo do ITU crítico superior. Diante da análise dos resultados, pôde-se observar em relação ao comportamento alimentar, que a fêmea Landrace apresentou o menor tempo em alimentação com médias de 42,19 min dia-1 e 43,73 min dia-1 para o inverno e verão, respectivamente, seguido do macho castrado de mesma linhagem. Enquanto as demais linhagens apresentaram valores acima de 60 min dia-1. Não foi observado correlação linear significativa entre o ambiente térmico e TM uma vez que os animais estiveram dentro de sua ZCT ao longo de todo período experimental, indicando que o comportamento alimentar foi influenciado principalmente pelos fatores homeostáticos e cognitivos-hedônicos. A estatística multivariada dividiu os animais em 8 grupos. Foi observado que animais de linhagens e sexos distintos se comportaram da mesma maneira, dificultando a modelagem matemática. Entretanto, alguns grupos apresentaram maior quantidade de animais de determinada linhagem e sexo, sendo estes utilizados como \"grupos padrão\" para o desenvolvimento do modelo fuzzy e a rede neural artificial. O modelo fuzzy apresentou R2 de 0,858 quando utilizado os dados do grupo padrão, entretanto, para todos os valores o R2 foi de 0,549. Já a rede neural apresentou um R2 de 0,611 para os dados completos e R2 de 0,914 para o \"grupo padrão\". Portanto, a rede neural artificial mostrou-se como uma ferramenta de maior precisão e acurácia na predição do comportamento alimentar de suínos nas fases de crescimento e terminação.The swine production in an activity of great importance to Brazil and to the world. However, because they maintain a constant body temperature and, alterations in the thermic accommodation environment can directly affect their physiological and behavioral responses for maintaining the internal temperature. Thus, the objective of this study was to access the feeding behavior of growing-finishing pigs of different sirelines and gender and its relationship with climate variables (thermic environment). Furthermore, mathematical models based on classic logic was developed as well as an intelligent system for predicting the total time spent eating (TM, min day -1). The data was collected in an experimental farm located in Clay Center, Nebraska, United States. The experimental period contemplated two seasons (summer and winter), totalizing 63 days (9 weeks) of information collected for each season. The housed animals were from three different commercial sirelines: Landrace, Duroc and Yorkshire. Each pen presented a mix composition, being housed 40 animals of different sirelines and gender. In total, there were 240 housed animals, being 80 animals for each sireline among barrows and gilts. The data registered were air temperature (Tar, °C), dew point temperature (Tpo, °C) and relative humidity of the air (UR, %) every 5 minutes inside the facility. For TM, the data were collected and registered every 20 seconds by a radio frequency data collection system. The thermal comfort was analyzed from the Temperature and Humidity Index (THI) and Specific Enthalpy (H, kJ kg-1 of dry air). In order to evaluate the relationship between the thermic environment and TM, the multivariate statistics through principal component analysis (PCA) and grouping was utilized for obtaining the selection standards of variables to enter in the models. The fuzzy model and the artificial neural networks were developed in a MATLAB® R2015a environment through the Fuzzy and the Neural Network toolboxes with the objective to predict TM, having as entry variables: sireline, gender, age and THI. On the whole, the Tar averages were inside the thermoneutral zone (ZCT), however, these values were below the superior critic THI. In the face of the results analysis, it could be observed in ration to the feeding behavior that the Landrace gilt presented the shortest time eating with averages of 42.19 min day-1 and 43.73 min day-1 for winter and summer respectively followed by the barrow from the same sireline, while the other sirelines presented values above 60 min day-1. It was not observed a significative linear correlation between the thermic environment and TM once the animals were inside their ZCT throughout all the experimentation period, indicating that the feeding behavior was influenced mainly by the homeostatic and cognitivehedonic factors. The multivariate statistics divided the animals in 8 groups, being observed that animals of different sirelines and gender behave the same way throughout the experimentation period, making the mathematical modeling difficult. However, some groups presented a bigger amount of animals of determined sireline and gender, being utilized as \"standard groups\" for the development of the fuzzy model and the artificial neural network. The fuzzy model presented an R2 of 0,858 when utilizing the \"standard group\" data, however, for all the values the R2 was 0.549. In the other hand the neural network presented an R2 of 0.611 for the complete data and an R2 of 0.914 for the \"standard group\". Thus, the artificial neural network appeared to be a tool of a better precision and accuracy when predicting the feeding behavior of pigs on growing-finishing phases.
6
Cambraia, Mario Sergio. "Automação da redução de perdas técnicas nos sistemas reticulados de distribuição utilizando redes neurais artificiais em redes inteligentes (smart grid)." Universidade de São Paulo, 2017. http://www.teses.usp.br/teses/disponiveis/3/3143/tde-05032018-102829/.
Full textAbstract:
Este trabalho apresenta a metodologia, o desenvolvimento e testes de um sistema de automação independente, baseado em Redes Neurais Artificiais, para redução de perdas técnicas em redes de distribuição subterrâneas reticuladas por meio do controle ótimo dos bancos de capacitores presentes na rede. A metodologia proposta contempla funcionalidades típicas de Redes Inteligentes, incluindo soluções práticas para o posicionamento de sensores de corrente em redes subterrâneas, coleta de medições de campo e transmissão para o Centro de Operação da Distribuição e controle em tempo real dos equipamentos de campo (bancos de capacitores). Portanto este trabalho consiste na implementação da solução através de baixo custo de investimento na mitigação do controle do fator de potência nos pontos de entrega ao consumidor, sendo que com isto ocorrem melhorias nos indicadores de qualidade e confiabilidade atendendo aos requisitos regulamentares e contratuais de fornecimento das distribuidoras. Para validação da metodologia proposta, foram utilizados os dados da concessionária de energia AES Eletropaulo sobre a Rede de Distribuição Subterrânea Reticulada do centro da cidade de São Paulo. As etapas da metodologia proposta e os principais aspectos do desenvolvimento do sistema são também descritos, bem como os testes realizados para comprovação dos resultados e validação do sistema.This work presents the methodology, development and testing of an independent automation system, based on Artificial Neural Networks, to reduce technical losses in reticulated underground distribution networks by means of the optimal control of the capacitor banks present in the network. The proposed methodology includes typical functionalities of Intelligent Networks, including practical solutions for the positioning of current sensors in underground networks, collection of field measurements and transmission to the Distribution Operation Center and real-time control of field equipment (capacitors banks). Therefore, this work consists in the implementation of the solution through a low cost of investment in the mitigation of the control of the power factor in the points of delivery to the consumer, and with this there are improvements in the indicators of quality and reliability taking into account the regulatory and contractual requirements of supply of the distributors. The energy concessionaire AES Eletropaulo had great participation in this research project, providing the necessary data of the Reticulated Underground Distribution Network of the city center of São Paulo. The steps of the proposed methodology and the main aspects of system development are also described, as well as the tests performed to prove the results and validate the system.
7
Moretti, José Fernando [UNESP]. "Sistema inteligente baseado nas redes neurais artificiais para dosagem do concreto." Universidade Estadual Paulista (UNESP), 2010. http://hdl.handle.net/11449/100325.
Full textAbstract:
Made available in DSpace on 2014-06-11T19:30:50Z (GMT). No. of bitstreams: 0
Previous issue date: 2010-10-05Bitstream added on 2014-06-13T18:40:59Z : No. of bitstreams: 1
moretti_jf_dr_ilha.pdf: 2346139 bytes, checksum: 4b4bfcffd24744e4627ebd26e46f3196 (MD5)Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
O concreto é o material estrutural mais utilizado na construção civil. Há mais de um século e meio ele vem sendo estudado e aperfeiçoado. É confeccionado utilizando-se de matérias primas regionais, com características técnicas diferentes de outras regiões. O cimento também se apresenta com diversas formulações. Quantificar adequadamente esses materiais é o objetivo do estudo da dosagem do concreto, de tal modo a se obter um concreto que atenda às necessidades estruturais exigidas. Sendo a principal delas a resistência à compressão. A dosagem do concreto é uma prática essencialmente laboratorial quando se pensa em resultados aceitáveis. Através de experimentos são idealizados ábacos e diagramas que podem fornecer a resistência do concreto endurecido com diversas combinações de matérias primas utilizadas. Não há uma formulação matemática abrangente e bem definida para um processo generalizado de dosagem. A complexidade aumenta quando se adicionam outros componentes a mais no concreto simples e tradicional. Obter a relação entre eles é um trabalho contínuo. As redes neurais vêm sendo utilizadas na solução de problemas da engenharia civil, com ênfase na aplicação da técnica da retropropagação. Ela realiza satisfatoriamente as iterações entre as diversas variáveis, num processo de treinamento e aprendizagem, e tem sido capaz de generalizar soluções aceitáveis. Nesta pesquisa de doutorado é utilizada uma rede neural feedfoward com algoritmo retropropagação para prever a resistência e o módulo de elasticidade do concreto. Os dados de entrada são quantidades de materiais utilizadas para confeccionar 1 m3 de concreto adensado, de forma semelhante a dosagem por diagramas. Será aplicada na interpretação de diagramas de dosagem. Tais diagramas são amplamente utilizados por empresas na confecção de concretos,...
Concrete is the most widely used structural material in construction, for more than a century and a half it has been studied and improved. It's prepared using regional raw materials with different technical characteristics of other regions. The cement also performs with various formulations. Adequately quantify these materials is the goal of the study of the concrete mixtures proportion, to obtain a concrete that meets the structural needs required. The main one being the compressive strength. The strength of concrete is essentially a practice laboratory when one considers acceptable results. Through experiments are idealized abacus and diagrams that can provide the strength of hardened concrete with various combinations of raw materials used. There is no mathematical formulation of comprehensive and well defined for a generalized process of mixes. The complexity increases when other components is added in the most simple and traditional concrete. Obtain the relationship between them is a work in progress. Neural networks have been used in solving engineering problems, with emphasis on applying the technique of backpropagation. It performs satisfactorily iterations between the different variables in a process of training and learning, and has been able to generalize acceptable solutions. In this work is used a feedforward neural network with backpropagation algorithm to predict the compressive strength and modulus of elasticity of the concrete. The input data are quantities of materials used to fabricate 1,0 m3 of concrete hardened, similar dosing for diagrams and abacus. Such diagrams are widely used by companies in the manufacturing of concrete, yielding good precision in the final results. They are produced on the vast experience with the same materials and are highly regional representative to provide subsidies for training neural networks. This... (Complete abstract click electronic access below)
8
Barbato, Daniela Maria Lemos. "O efeito das lesões nas capacidades de memorização e generalização de um perceptron." Universidade de São Paulo, 1993. http://www.teses.usp.br/teses/disponiveis/54/54131/tde-05092008-144618/.
Full textAbstract:
Perceptrons são redes neurais sem retroalimentação onde os neurônios estão dispostos em camadas. O perceptron considerado neste trabalho consiste de uma camada de N neurônios sensores Si = ±1; i = 1, , N ligados a um neurônio motor δ através das conexões sinápticas (pesos) Wi; i = 1, ..., N cujos valores restringimos a ±1. Utilizando o formalismo de Mecânica Estatística desenvolvido por Gardner (1988), estudamos os efeitos de eliminarmos uma fração de conexões sinápticas (diluição ) nas capacidades de memorização e generalização da rede neural descrita acima. Consideramos também o efeito de ruído atuando durante o estágio de treinamento do perceptron. Consideramos dois tipos de diluição: diluição móvel na qual os pesos são cortados de maneira a minimizar o erro de treinamento e diluição fixa na qual os pesos são cortados aleatoriamente. A diluição móvel, que modela lesões em cérebro de pacientes muito jovens, pode melhorar a capacidade de memorização e, no caso da rede ser treinada com ruído, também pode melhorar a capacidade de generalização. Por outro lado, a diluição fixa, que modela lesões em cérebros de pacientes adultos, sempre degrada o desempenho da rede, sendo seu principal efeito introduzir um ruído efetivo nos exemplos de treinamento.Perceptrons are layered, feed-forward neural networks. In this work we consider a per-ceptron composed of one input layer with N sensor neurons Si = ±1; i = 1, ... , N which are connected to a single motor neuron δ through the synaptic weights Wj; i = 1, ... , N, which are constrained to take on the values ±1 only. Using the Statistical Mechanics formalism developed by Gardner (1988), we study the effects of eliminating a fraction of synaptic weights on the memorization and generalization capabilities of the neural network described above. We consider also the effects of noise acting during the perceptron training stage. We consider two types of dilution: annealed dilution, where the weights are cut so as to minimize the training error and quenched dilution, where the weights are cut randomly. The annealed dilution which models brain damage in very young patients can improve the memorization ability and, in the case of training with noise, it can also improve the generalization ability. On the other hand, the quenched dilution which models lesions on adult brains always degrades the performance of the network, its main effect being to introduce an effective noise in the training examples.
9
Janes, Ricardo. "Proposição de um algoritmo para identificação biométrica de pessoas baseado nos padrões de veias das mãos." Universidade de São Paulo, 2015. http://www.teses.usp.br/teses/disponiveis/3/3143/tde-20072016-082931/.
Full textAbstract:
Esta tese tem por finalidade apresentar o desenvolvimento de um sistema biométrico de baixo custo, capaz de identificar pessoas pela análise dos padrões de veias das mãos com obtenção de imagens no espectro infravermelho próximo. O sistema foi montado fisicamente através da construção de um protótipo e então foram aquisitadas e armazenadas 520 imagens da parte dorsal da mão direita de 52 diferentes usuários, após isto foi realizada a extração de uma região de interesse definida pela maior porção quadrada da parte dorsal da mão. Em seguida foram aplicados três diferentes métodos de equalização e suavização da imagem na fase de pré-processamento, para posterior extração das características das veias com a utilização da transformada de Curvelet na função \"wrapping\" e aplicação do algoritmo Padrão Binário Local (LBP) para a digitalização do conteúdo extraído. No próximo passo, uma análise de identificação foi realizada usando cinco diferentes métodos de classificação. Em primeiro lugar, foi utilizado um classificador probabilístico Naive Bayes, em seguida um classificador baseado em aprendizagem por regressão linear Kernel Nearest Neighbor (K-NN), ainda foram aplicados dois algoritmos baseados em árvores de decisão C4.5 e Random Forest e finalmente um algoritmo baseado em redes neurais artificiais Multilayer Perceptron. Os classificadores foram testados utilizando o método de validação cruzada, e as informações foram separadas por 10 folds sendo que 10% dos dados foram utilizados para treino e 90% dos dados foram utilizados para teste. Com os mesmos dados resultantes da fase de pré-processamento, dois algoritmos foram aplicados para seleção de características, sendo o primeiro baseado na correlação da função de seleção de recursos e o segundo na seleção de atributos pelo conceito da entropia dos dados. Os resultados provam que o método de equalização de histograma adaptativa por limite de contraste na fase de pré-processamento apresentou os melhores resultados. Quanto aos classificadores, os melhores resultados foram obtidos com o uso da rede neural artificial proposta e as taxas de falsa aceitação (FAR) e falsa rejeição (FRR) obtidas após o processamento foram estimadas em 0,038 e 0,003 respectivamente. Foram realizados ainda testes com a quantidade mínima de imagens necessárias para identificação de pessoas e chegou-se ao valor de cinco imagens por usuário. Finalmente a avaliação da permanência do sistema biométrico foi realizada através da análise de imagens capturadas após um ano da primeira análise e os resultados mostram que o sistema é robusto, apesar das imagens conterem pequenas alterações, proporcionais às variações do índice de massa corporal dos usuários.The system has been assembled as a prototype then were acquired and storaged 520 images from the dorsal side of the right hand of 52 different users, and then is accomplished an extracting of a region of interest defined by the largest square portion of the dorsal hand. Then a pre-processing of image has been applied using three different methods of image equalization and smoothing for later extraction of the veins characteristics using the Curvelet Transform in \"wrapping\" function and application of the Local Binary Pattern algorithm (LBP) for scanning the extracted content. On the next step, an identification analysis has been performed using five different classification methods. First, a probabilistic Naive Bayes classifier was used, second a classifier based on linear regression called Kernel Nearest Neighbor (K-NN) was applied, third and fourth two algorithms based on decision trees, C4.5 and Random Forest were tested, and finally an algorithm based on artificial neural networks Multilayer Perceptron was performed. The classifiers have been tested using the cross-validation method, and the information was separated by 10 folds wherein 10% of the data were used for training and 90% of the data were used for testing. From the same data resulted of the pre-processing step, two algorithms have been applied for selection features, the first based on the correlation based feature selection and the second in selecting attributes based to the concept of entropy data. The results proof that the equalization method by contrast limited adaptive histogram equalization, in the pre-processing stage, shown the best results. From the application of classifiers, the best result was achieved by using the artificial neural network proposal and the false acceptance rate (FAR) and false rejection rate (FRR) found through the processing were estimated in 0.038 and 0.003 respectively. Tests were also performed to assess the minimum amount of images needed to identify people and as result five images per user were found as the ideal number. Finally, the assessment of the biometric system permanence was performed using acquired images after a year of the first analysis and the results shown that the system is robust, even that the pictures contain minor changes proportional to index variations of body mass of users.
10
Mon-Ma, Marly Mitiko. "Análise da importância das variáveis intervenientes nos acidentes de trânsito em interseções urbanas utilizando redes neurais artificiais." Universidade Federal de São Carlos, 2005. https://repositorio.ufscar.br/handle/ufscar/4403.
Full textAbstract:
Made available in DSpace on 2016-06-02T20:01:03Z (GMT). No. of bitstreams: 1
DissMMM.pdf: 1370681 bytes, checksum: 4891da17b25caf7951985f7209d9c825 (MD5)
Previous issue date: 2005-10-06The technological development has generated great amount of potential data bases in order to supply information for several aspects related to road safety. However, the transformation of these great amount of data in useful information for the technicians, public managers and the population in general, requests the modeling and the treatment of these data using some analysis tools that allow a visualization of the results in form easily understandable. This work presents a new methodology of traffic accidents analysis based in the artificial neural network (ANN). ANN can be very useful for organizations, public or particular, mainly to those that propose to understand the phenomena of the traffic in order to looking for solutions integrated to several areas such as education, engineering and fiscalization. This research had as general objective to identify the patterns of traffic accidents that happened at urban intersections. The data of accidents that happened in the period from 2000 to 2003, in the city of São Carlos were used for the case study, in order to subsidize the elaboration and the evaluation of public policies of traffic accidents reduction and specially the reduction of accident severity. The study explores the assumption that different accident types are related to different patterns. The patterns obtained by ANN showed that there are significant differences in the factors that can affect the different types of accidents. The knowledge of the patterns of each accident type is essential to develop actions corrective or preventive road safety's improvement in order to avoid undesirable effects when these actions are implemented. However, the comparison between the patterns of the different types of accidents was difficult due to the heterogeneity of the situations and the different elements that compose the road environment that can affect the occurrence of the accident.
O desenvolvimento tecnológico tem gerado grandes bases de dados com potencial para fornecer informações sobre diversos aspectos relacionados com a segurança viária. No entanto, a conversão de um grande volume de dados em informações úteis para os técnicos, gestores públicos e a população em geral, requer a modelagem e o tratamento destes dados utilizando ferramentas de análise que permitam uma visualização dos resultados de forma facilmente compreensível. Este trabalho apresenta uma nova metodologia para análise de acidentes de trânsito fundamentada na rede neural artificial (RNA). A RNA pode ser de grande utilidade para organizações públicas e privadas, principalmente para aquelas que se propõem compreender os fenômenos do trânsito a fim de buscar soluções integradas em diversas áreas tais como educação, engenharia e fiscalização. A pesquisa teve como objetivo geral identificar os padrões de acidentes de trânsito que ocorreram nas interseções urbanas. Os dados de acidentes que ocorreram no período de 2000 a 2003, na cidade de São Carlos foram utilizados para o estudo de caso, visando fornecer subsídios para a elaboração e a avaliação de políticas públicas voltadas para redução do número de acidentes de trânsito e essencialmente na redução global da severidade. O estudo explora a suposição de que diferentes tipos de acidente estão relacionados com padrões distintos. Os padrões obtidos através da RNA mostram que há divergências significativas nos fatores que podem influenciar os diferentes tipos de acidentes. Conhecer padrões de cada tipo de acidente se faz necessária para que as medidas corretivas ou preventivas voltadas para a melhoria da segurança viária não resultem em efeitos indesejados quando são implementadas, no entanto comparações entre padrões de diferentes tipos de acidentes mostraram-se particularmente difíceis devido à heterogeneidade das situações e dos diferentes elementos que compõem o ambiente viário e que podem influenciar na ocorrência do acidente.
Books on the topic "Neural networks (NNs)"
1
Raff, Lionel, Ranga Komanduri, Martin Hagan, and Satish Bukkapatnam. Neural Networks in Chemical Reaction Dynamics. Oxford University Press, 2012. http://dx.doi.org/10.1093/oso/9780199765652.001.0001.
Full textAbstract:
This monograph presents recent advances in neural network (NN) approaches and applications to chemical reaction dynamics. Topics covered include: (i) the development of ab initio potential-energy surfaces (PES) for complex multichannel systems using modified novelty sampling and feedforward NNs; (ii) methods for sampling the configuration space of critical importance, such as trajectory and novelty sampling methods and gradient fitting methods; (iii) parametrization of interatomic potential functions using a genetic algorithm accelerated with a NN; (iv) parametrization of analytic interatomic potential functions using NNs; (v) self-starting methods for obtaining analytic PES from ab inito electronic structure calculations using direct dynamics; (vi) development of a novel method, namely, combined function derivative approximation (CFDA) for simultaneous fitting of a PES and its corresponding force fields using feedforward neural networks; (vii) development of generalized PES using many-body expansions, NNs, and moiety energy approximations; (viii) NN methods for data analysis, reaction probabilities, and statistical error reduction in chemical reaction dynamics; (ix) accurate prediction of higher-level electronic structure energies (e.g. MP4 or higher) for large databases using NNs, lower-level (Hartree-Fock) energies, and small subsets of the higher-energy database; and finally (x) illustrative examples of NN applications to chemical reaction dynamics of increasing complexity starting from simple near equilibrium structures (vibrational state studies) to more complex non-adiabatic reactions. The monograph is prepared by an interdisciplinary group of researchers working as a team for nearly two decades at Oklahoma State University, Stillwater, OK with expertise in gas phase reaction dynamics; neural networks; various aspects of MD and Monte Carlo (MC) simulations of nanometric cutting, tribology, and material properties at nanoscale; scaling laws from atomistic to continuum; and neural networks applications to chemical reaction dynamics. It is anticipated that this emerging field of NN in chemical reaction dynamics will play an increasingly important role in MD, MC, and quantum mechanical studies in the years to come.
2
Fox, Raymond. The Use of Self. Oxford University Press, 2011. http://dx.doi.org/10.1093/oso/9780190616144.001.0001.
Full textAbstract:
This monograph presents recent advances in neural network (NN) approaches and applications to chemical reaction dynamics. Topics covered include: (i) the development of ab initio potential-energy surfaces (PES) for complex multichannel systems using modified novelty sampling and feedforward NNs; (ii) methods for sampling the configuration space of critical importance, such as trajectory and novelty sampling methods and gradient fitting methods; (iii) parametrization of interatomic potential functions using a genetic algorithm accelerated with a NN; (iv) parametrization of analytic interatomic potential functions using NNs; (v) self-starting methods for obtaining analytic PES from ab inito electronic structure calculations using direct dynamics; (vi) development of a novel method, namely, combined function derivative approximation (CFDA) for simultaneous fitting of a PES and its corresponding force fields using feedforward neural networks; (vii) development of generalized PES using many-body expansions, NNs, and moiety energy approximations; (viii) NN methods for data analysis, reaction probabilities, and statistical error reduction in chemical reaction dynamics; (ix) accurate prediction of higher-level electronic structure energies (e.g. MP4 or higher) for large databases using NNs, lower-level (Hartree-Fock) energies, and small subsets of the higher-energy database; and finally (x) illustrative examples of NN applications to chemical reaction dynamics of increasing complexity starting from simple near equilibrium structures (vibrational state studies) to more complex non-adiabatic reactions. The monograph is prepared by an interdisciplinary group of researchers working as a team for nearly two decades at Oklahoma State University, Stillwater, OK with expertise in gas phase reaction dynamics; neural networks; various aspects of MD and Monte Carlo (MC) simulations of nanometric cutting, tribology, and material properties at nanoscale; scaling laws from atomistic to continuum; and neural networks applications to chemical reaction dynamics. It is anticipated that this emerging field of NN in chemical reaction dynamics will play an increasingly important role in MD, MC, and quantum mechanical studies in the years to come.
3
Padilla, Claudia R., and Mario F. Mendez. Neuropsychiatric Features Across Neurodegenerative Diseases. Oxford University Press, 2016. http://dx.doi.org/10.1093/med/9780190233563.003.0006.
Full textAbstract:
Neuropsychiatric symptoms (NPS) are a major manifestation of neurodegenerative diseases(NDDs) including Alzheimer’s disease (AD), Dementia with Lewy Bodies (DLB) and frontotemporal dementia (FTD). NPS symptoms are a determining factor impacting economic and psychological costs for both the patient and their caregivers in these devastating illness. Recent developments in neuroscience have clarified the relationship of NPS with changes in brain structures, alterations in neural circuits and networks, and their neurotransmitter systems. It is increasingly recognized that NPS shared across different NDDs might have shared alterations in neural circuits and networks, and their neurotransmitter systems and ways to modulate them theraputcally. This chapter focuses on the more common NPS and their links to causative neurodegenerative processes that transcend specific diseases.
Book chapters on the topic "Neural networks (NNs)"
1
Müller, Achim F., and Hans Georg Zimmermann. "Symbolic Prosody Modeling by Causal Retro-causal NNs with Variable Context Length." In Artificial Neural Networks — ICANN 2001, 57–64. Berlin, Heidelberg: Springer Berlin Heidelberg, 2001. http://dx.doi.org/10.1007/3-540-44668-0_9.
Full text
2
Dong, Guowei, Yongming Li, Duo Meng, Fuming Sun, and Rui Bai. "Adaptive NNs Fault-Tolerant Control for Nonstrict-Feedback Nonlinear Systems." In Advances in Neural Networks - ISNN 2017, 11–19. Cham: Springer International Publishing, 2017. http://dx.doi.org/10.1007/978-3-319-59081-3_2.
Full text
3
Khedr, Haitham, James Ferlez, and Yasser Shoukry. "PEREGRiNN: Penalized-Relaxation Greedy Neural Network Verifier." In Computer Aided Verification, 287–300. Cham: Springer International Publishing, 2021. http://dx.doi.org/10.1007/978-3-030-81685-8_13.
Full textAbstract:
AbstractNeural Networks (NNs) have increasingly apparent safety implications commensurate with their proliferation in real-world applications: both unanticipated as well as adversarial misclassifications can result in fatal outcomes. As a consequence, techniques of formal verification have been recognized as crucial to the design and deployment of safe NNs. In this paper, we introduce a new approach to formally verify the most commonly considered safety specifications for ReLU NNs – i.e. polytopic specifications on the input and output of the network. Like some other approaches, ours uses a relaxed convex program to mitigate the combinatorial complexity of the problem. However, unique in our approach is the way we use a convex solver not only as a linear feasibility checker, but also as a means of penalizing the amount of relaxation allowed in solutions. In particular, we encode each ReLU by means of the usual linear constraints, and combine this with a convex objective function that penalizes the discrepancy between the output of each neuron and its relaxation. This convex function is further structured to force the largest relaxations to appear closest to the input layer; this provides the further benefit that the most “problematic” neurons are conditioned as early as possible, when conditioning layer by layer. This paradigm can be leveraged to create a verification algorithm that is not only faster in general than competing approaches, but is also able to verify considerably more safety properties; we evaluated PEREGRiNN on a standard MNIST robustness verification suite to substantiate these claims.
4
Kim, Wook-Dong, Sung-Kwun Oh, and Hyun-Ki Kim. "Fuzzy Clustering-Based Polynomial Radial Basis Function Neural Networks (p-RBF NNs) Classifier Designed with Particle Swarm Optimization." In Advances in Neural Networks – ISNN 2011, 464–73. Berlin, Heidelberg: Springer Berlin Heidelberg, 2011. http://dx.doi.org/10.1007/978-3-642-21105-8_54.
Full text
5
Mamalakis, Antonios, Imme Ebert-Uphoff, and Elizabeth A. Barnes. "Explainable Artificial Intelligence in Meteorology and Climate Science: Model Fine-Tuning, Calibrating Trust and Learning New Science." In xxAI - Beyond Explainable AI, 315–39. Cham: Springer International Publishing, 2022. http://dx.doi.org/10.1007/978-3-031-04083-2_16.
Full textAbstract:
AbstractIn recent years, artificial intelligence and specifically artificial neural networks (NNs) have shown great success in solving complex, nonlinear problems in earth sciences. Despite their success, the strategies upon which NNs make decisions are hard to decipher, which prevents scientists from interpreting and building trust in the NN predictions; a highly desired and necessary condition for the further use and exploitation of NNs’ potential. Thus, a variety of methods have been recently introduced with the aim of attributing the NN predictions to specific features in the input space and explaining their strategy. The so-called eXplainable Artificial Intelligence (XAI) is already seeing great application in a plethora of fields, offering promising results and insights about the decision strategies of NNs. Here, we provide an overview of the most recent work from our group, applying XAI to meteorology and climate science. Specifically, we present results from satellite applications that include weather phenomena identification and image to image translation, applications to climate prediction at subseasonal to decadal timescales, and detection of forced climatic changes and anthropogenic footprint. We also summarize a recently introduced synthetic benchmark dataset that can be used to improve our understanding of different XAI methods and introduce objectivity into the assessment of their fidelity. With this overview, we aim to illustrate how gaining accurate insights about the NN decision strategy can help climate scientists and meteorologists improve practices in fine-tuning model architectures, calibrating trust in climate and weather prediction and attribution, and learning new science.
6
Baniadamdizaj, Shima. "Localization Using DeepLab in Document Images Taken by Smartphones." In Digital Interaction and Machine Intelligence, 63–74. Cham: Springer International Publishing, 2022. http://dx.doi.org/10.1007/978-3-031-11432-8_6.
Full textAbstract:
AbstractThe seamless integration of statistics from virtual and paper files could be very crucial for the know-how control of efficient. A handy manner to obtain that is to digitize a report from a picture. This calls for the localization of the report in the picture. Several approaches are deliberate to resolve this hassle; however, they are supported historical picture method strategies that are not robust to intense viewpoints and backgrounds. Deep Convolutional Neural Networks (CNNs), on the opposite hand, have been validated to be extraordinarily strong to versions in heritage and perspective of view for item detection and classification duties. Inspired by their robustness and generality, we advocate a CNN-primarily based totally technique for the correct localization of files in real-time. We advocate the new utilization of Neural Networks (NNs) for the localization hassle as a key factor detection hassle. The proposed technique ought to even localize snapshots that don't have a very square shape. Also, we used a newly amassed dataset that has extra tough duties internal and is in the direction of a slipshod user. The result is knowledgeable in 3 specific classes of snapshots and our proposed technique has 100% accuracy on easy one and 77% on average. The result is as compared with the maximum famous report localization strategies and cell applications.
7
Aoun, Mario Antoine. "STDP within NDS Neurons." In Advances in Neural Networks - ISNN 2010, 33–43. Berlin, Heidelberg: Springer Berlin Heidelberg, 2010. http://dx.doi.org/10.1007/978-3-642-13278-0_5.
Full text
8
Gridin, Ivan. "NNI Recipes." In Automated Deep Learning Using Neural Network Intelligence, 357–77. Berkeley, CA: Apress, 2022. http://dx.doi.org/10.1007/978-1-4842-8149-9_7.
Full text
9
Ernesto, Burattini. "NES: a Neuron-like net for a diagnostic Expert System." In International Neural Network Conference, 675. Dordrecht: Springer Netherlands, 1990. http://dx.doi.org/10.1007/978-94-009-0643-3_41.
Full text
10
Tran, Hoang-Dung, Neelanjana Pal, Patrick Musau, Diego Manzanas Lopez, Nathaniel Hamilton, Xiaodong Yang, Stanley Bak, and Taylor T. Johnson. "Robustness Verification of Semantic Segmentation Neural Networks Using Relaxed Reachability." In Computer Aided Verification, 263–86. Cham: Springer International Publishing, 2021. http://dx.doi.org/10.1007/978-3-030-81685-8_12.
Full textAbstract:
AbstractThis paper introduces robustness verification for semantic segmentation neural networks (in short, semantic segmentation networks [SSNs]), building on and extending recent approaches for robustness verification of image classification neural networks. Despite recent progress in developing verification methods for specifications such as local adversarial robustness in deep neural networks (DNNs) in terms of scalability, precision, and applicability to different network architectures, layers, and activation functions, robustness verification of semantic segmentation has not yet been considered. We address this limitation by developing and applying new robustness analysis methods for several segmentation neural network architectures, specifically by addressing reachability analysis of up-sampling layers, such as transposed convolution and dilated convolution. We consider several definitions of robustness for segmentation, such as the percentage of pixels in the output that can be proven robust under different adversarial perturbations, and a robust variant of intersection-over-union (IoU), the typical performance evaluation measure for segmentation tasks. Our approach is based on a new relaxed reachability method, allowing users to select the percentage of a number of linear programming problems (LPs) to solve when constructing the reachable set, through a relaxation factor percentage. The approach is implemented within NNV, then applied and evaluated on segmentation datasets, such as a multi-digit variant of MNIST known as M2NIST. Thorough experiments show that by using transposed convolution for up-sampling and average-pooling for down-sampling, combined with minimizing the number of ReLU layers in the SSNs, we can obtain SSNs with not only high accuracy (IoU), but also that are more robust to adversarial attacks and amenable to verification. Additionally, using our new relaxed reachability method, we can significantly reduce the verification time for neural networks whose ReLU layers dominate the total analysis time, even in classification tasks.
Conference papers on the topic "Neural networks (NNs)"
1
Benussi, Elias, Andrea Patane', Matthew Wicker, Luca Laurenti, and Marta Kwiatkowska. "Individual Fairness Guarantees for Neural Networks." In Thirty-First International Joint Conference on Artificial Intelligence {IJCAI-22}. California: International Joint Conferences on Artificial Intelligence Organization, 2022. http://dx.doi.org/10.24963/ijcai.2022/92.
Full textAbstract:
We consider the problem of certifying the individual fairness (IF) of feed-forward neural networks (NNs). In particular, we work with the epsilon-delta-IF formulation, which, given a NN and a similarity metric learnt from data, requires that the output difference between any pair of epsilon-similar individuals is bounded by a maximum decision tolerance delta >= 0. Working with a range of metrics, including the Mahalanobis distance, we propose a method to overapproximate the resulting optimisation problem using piecewise-linear functions to lower and upper bound the NN's non-linearities globally over the input space. We encode this computation as the solution of a Mixed-Integer Linear Programming problem and demonstrate that it can be used to compute IF guarantees on four datasets widely used for fairness benchmarking. We show how this formulation can be used to encourage models' fairness at training time by modifying the NN loss, and empirically confirm our approach yields NNs that are orders of magnitude fairer than state-of-the-art methods.
2
Pollmeier, K., C. R. Burrows, and K. A. Edge. "Condition Monitoring of an Electrohydraulic Position Control System Using Artificial Neural Networks." In ASME 2004 International Mechanical Engineering Congress and Exposition. ASMEDC, 2004. http://dx.doi.org/10.1115/imece2004-62309.
Full textAbstract:
This paper investigates the condition monitoring of a servo-valve-controlled linear actuator system using artificial neural networks (NNs). The aim is to discuss techniques for the identification of failure characteristics and their source. It is shown that neural networks can be trained to identify more than one fault but these are larger and require more training patterns than networks for single fault diagnosis. This leads to much longer training times and to problems with scaleability. Therefore a modular approach has been developed. Several networks were trained each to identify an individual fault. The parallel outputs of these nets were then used as inputs to another network. This additional network was able to identify not only the correct faults but also the actual fault levels.
3
Ghoreyshi, M., P. Pilidis, and K. W. Ramsden. "Diagnostics of a Small Jet Engine-Neural Networks Approach." In ASME Turbo Expo 2005: Power for Land, Sea, and Air. ASMEDC, 2005. http://dx.doi.org/10.1115/gt2005-68511.
Full textAbstract:
This paper presents the design procedure and application of a nested neural network for diagnostics of a small jet engine. Such a diagnostics technique is based on the performance analysis while the performance model was developed with TURBOMATCH, the Cranfield University’s gas turbine simulation code. To validate this model, an outdoor test was conducted to run the small engine. Areas examined in this paper are performance validation of the engine, neural network design, training data generation, and networks training procedures. The assumptions, measured parameters selection and the results obtained are presented and discussed. The results obtained show the good prospects for the use of NNs for detection of existing faults, isolation of faults and quantification of fault levels.
4
Zhan, Huixin, Kun Zhang, Chenyi Hu, and Victor S. Sheng. "Gated Graph Neural Networks (GG-NNs) for Abstractive Multi-Comment Summarization." In 2021 IEEE International Conference on Big Knowledge (ICBK). IEEE, 2021. http://dx.doi.org/10.1109/ickg52313.2021.00050.
Full text
5
Szabo, T., L. Antoni, G. Horvath, and B. Feher. "A full-parallel digital implementation for pre-trained NNs." In Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks. IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium. IEEE, 2000. http://dx.doi.org/10.1109/ijcnn.2000.857873.
Full text
6
Wang, Jun, Kevin Chiu, and Mark Fuge. "Learning to Abstract and Compose Mechanical Device Function and Behavior." In ASME 2020 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. American Society of Mechanical Engineers, 2020. http://dx.doi.org/10.1115/detc2020-22714.
Full textAbstract:
Abstract While current neural networks (NNs) are becoming good at deriving single types of abstractions for a small set of phenomena, for example, using a single NN to predict a flow velocity field, NNs are not good at composing large systems as compositions of small phenomena and reasoning about their interactions. We want to study how NNs build both the abstraction and composition of phenomena when a single NN model cannot suffice. Rather than a single NN that learns one physical or social phenomenon, we want a group of NNs that learn to abstract, compose, reason, and correct the behaviors of different parts in a system. In this paper, we investigate the joint use of Physics-Informed (Navier-Stokes equations) Deep Neural Networks (i.e., Deconvolutional Neural Networks) as well as Geometric Deep Learning (i.e., Graph Neural Networks) to learn and compose fluid component behavior. Our models successfully predict the fluid flows and their composition behaviors (i.e., velocity fields) with an accuracy of about 99%.
7
Bettocchi, R., M. Pinelli, P. R. Spina, M. Venturini, and G. A. Zanetta. "Assessment of the Robustness of Gas Turbine Diagnostics Tools Based on Neural Networks." In ASME Turbo Expo 2006: Power for Land, Sea, and Air. ASMEDC, 2006. http://dx.doi.org/10.1115/gt2006-90118.
Full textAbstract:
The paper deals with the set-up and the application of an Artificial Intelligence technique based on Neural Networks (NNs) to gas turbine diagnostics, in order to evaluate its capabilities and its robustness. The data used for both training and testing the NNs were generated by means of a Cycle Program, calibrated on a Siemens V94.3A gas turbine. Such data are representative of operating points characterized by different boundary, load and health state conditions. The analyses carried out are aimed at the selection of the most appropriate NN structure for gas turbine diagnostics, by evaluating NN robustness with respect to: • interpolation capability and accuracy in the presence of data affected by measurement errors; • extrapolation capability in the presence of data lying outside the range of variation adopted for NN training; • accuracy in the presence of input data corrupted by bias errors; • accuracy when one input is not available. This situation is simulated by replacing the value of the unavailable input with its nominal value.
8
Haridas, Akash, and Nagabhushana Rao Vadlamani. "Modelling Wall-Pressure Spectra in Turbulent Boundary Layers Using Neural Networks." In ASME 2021 Gas Turbine India Conference. American Society of Mechanical Engineers, 2021. http://dx.doi.org/10.1115/gtindia2021-76301.
Full textAbstract:
Abstract In this work, we model the spectra of wall-pressure fluctuations beneath subsonic, supersonic and hypersonic turbulent boundary layers (TBLs) at zero pressure gradient using neural networks (NNs). We collect and compile data pertaining to wall-pressure fluctuation spectra from several experimental and computational studies on TBLs. In contrast to conventional methods of hand-tuning the parameters of a model to fit the available data, the use of modern powerful statistical learning techniques such as neural networks provide an automatic and quick way to fit a model. We explore four different scenarios of making use of the compiled data. In comparison with COMPRA-G, an empirical model recently proposed to account for compressibility effects in TBLs, we achieve a better fit to observed data using the NN model, particularly at low frequencies of the spectra.
9
Barclay, Andrew, and Jonathan Corney. "Automated Classification of Components for Manufacturing Planning: Single-View Convolutional Neural Network for Global Shape Identification." In ASME 2020 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. American Society of Mechanical Engineers, 2020. http://dx.doi.org/10.1115/detc2020-22335.
Full textAbstract:
Abstract An experienced engineer can glance at a component and suggest appropriate methods for its manufacture. This skill has been difficult to automate but in recent years Neural Networks have demonstrated impressive image recognition capabilities in many applications. Consequently, this work is motivated by the goal of automating shape assessment for manufacturing. Specifically the reported work investigates the feasibility of training a convolutional neural network (CNN) to recognize 2D images of shapes associated with particular Near Net Shape (NNS) manufacturing processes such as casting, forging, or flow forming. The system uses multiple images generated from 3D CAD models (each manually associated with specific NNS processes) as training data and a single shop floor photograph as a classification query. While multiple views are used to train the CNN only a single view is used to assess the accuracy of the classification. Such single-view classification is designed to support the easy assessment of physical parts observed in manufacturing facilities where it would often be impractical to create an array of images from many viewpoints. The result suggests that despite limitations, single-view CNNs can classify real engineering components for manufacture.
10
He, S., and N. Sepehri. "Experimental Study of a Neural Generalized Predictive Force Control for a Hydraulic Actuator." In ASME 2000 International Mechanical Engineering Congress and Exposition. American Society of Mechanical Engineers, 2000. http://dx.doi.org/10.1115/imece2000-2313.
Full textAbstract:
Abstract In this paper, multilayer feedforward neural networks (NNs) are used for modeling and force control of a hydraulic actuator. The predictability of the instantaneous linearized neural model is examined and is used along with the generalized predictive control (GPC) algorithm to control the force exerted on the environment. Experimental results show that the neural-based generalized predictive control can handle different contact environments despite high nonlinearity and uncertainty in the hydraulic functions.
Reports on the topic "Neural networks (NNs)"
1
Galili, Naftali, Roger P. Rohrbach, Itzhak Shmulevich, Yoram Fuchs, and Giora Zauberman. Non-Destructive Quality Sensing of High-Value Agricultural Commodities Through Response Analysis. United States Department of Agriculture, October 1994. http://dx.doi.org/10.32747/1994.7570549.bard.
Full textAbstract:
The objectives of this project were to develop nondestructive methods for detection of internal properties and firmness of fruits and vegetables. One method was based on a soft piezoelectric film transducer developed in the Technion, for analysis of fruit response to low-energy excitation. The second method was a dot-matrix piezoelectric transducer of North Carolina State University, developed for contact-pressure analysis of fruit during impact. Two research teams, one in Israel and the other in North Carolina, coordinated their research effort according to the specific objectives of the project, to develop and apply the two complementary methods for quality control of agricultural commodities. In Israel: An improved firmness testing system was developed and tested with tropical fruits. The new system included an instrumented fruit-bed of three flexible piezoelectric sensors and miniature electromagnetic hammers, which served as fruit support and low-energy excitation device, respectively. Resonant frequencies were detected for determination of firmness index. Two new acoustic parameters were developed for evaluation of fruit firmness and maturity: a dumping-ratio and a centeroid of the frequency response. Experiments were performed with avocado and mango fruits. The internal damping ratio, which may indicate fruit ripeness, increased monotonically with time, while resonant frequencies and firmness indices decreased with time. Fruit samples were tested daily by destructive penetration test. A fairy high correlation was found in tropical fruits between the penetration force and the new acoustic parameters; a lower correlation was found between this parameter and the conventional firmness index. Improved table-top firmness testing units, Firmalon, with data-logging system and on-line data analysis capacity have been built. The new device was used for the full-scale experiments in the next two years, ahead of the original program and BARD timetable. Close cooperation was initiated with local industry for development of both off-line and on-line sorting and quality control of more agricultural commodities. Firmalon units were produced and operated in major packaging houses in Israel, Belgium and Washington State, on mango and avocado, apples, pears, tomatoes, melons and some other fruits, to gain field experience with the new method. The accumulated experimental data from all these activities is still analyzed, to improve firmness sorting criteria and shelf-life predicting curves for the different fruits. The test program in commercial CA storage facilities in Washington State included seven apple varieties: Fuji, Braeburn, Gala, Granny Smith, Jonagold, Red Delicious, Golden Delicious, and D'Anjou pear variety. FI master-curves could be developed for the Braeburn, Gala, Granny Smith and Jonagold apples. These fruits showed a steady ripening process during the test period. Yet, more work should be conducted to reduce scattering of the data and to determine the confidence limits of the method. Nearly constant FI in Red Delicious and the fluctuations of FI in the Fuji apples should be re-examined. Three sets of experiment were performed with Flandria tomatoes. Despite the complex structure of the tomatoes, the acoustic method could be used for firmness evaluation and to follow the ripening evolution with time. Close agreement was achieved between the auction expert evaluation and that of the nondestructive acoustic test, where firmness index of 4.0 and more indicated grade-A tomatoes. More work is performed to refine the sorting algorithm and to develop a general ripening scale for automatic grading of tomatoes for the fresh fruit market. Galia melons were tested in Israel, in simulated export conditions. It was concluded that the Firmalon is capable of detecting the ripening of melons nondestructively, and sorted out the defective fruits from the export shipment. The cooperation with local industry resulted in development of automatic on-line prototype of the acoustic sensor, that may be incorporated with the export quality control system for melons. More interesting is the development of the remote firmness sensing method for sealed CA cool-rooms, where most of the full-year fruit yield in stored for off-season consumption. Hundreds of ripening monitor systems have been installed in major fruit storage facilities, and being evaluated now by the consumers. If successful, the new method may cause a major change in long-term fruit storage technology. More uses of the acoustic test method have been considered, for monitoring fruit maturity and harvest time, testing fruit samples or each individual fruit when entering the storage facilities, packaging house and auction, and in the supermarket. This approach may result in a full line of equipment for nondestructive quality control of fruits and vegetables, from the orchard or the greenhouse, through the entire sorting, grading and storage process, up to the consumer table. The developed technology offers a tool to determine the maturity of the fruits nondestructively by monitoring their acoustic response to mechanical impulse on the tree. A special device was built and preliminary tested in mango fruit. More development is needed to develop a portable, hand operated sensing method for this purpose. In North Carolina: Analysis method based on an Auto-Regressive (AR) model was developed for detecting the first resonance of fruit from their response to mechanical impulse. The algorithm included a routine that detects the first resonant frequency from as many sensors as possible. Experiments on Red Delicious apples were performed and their firmness was determined. The AR method allowed the detection of the first resonance. The method could be fast enough to be utilized in a real time sorting machine. Yet, further study is needed to look for improvement of the search algorithm of the methods. An impact contact-pressure measurement system and Neural Network (NN) identification method were developed to investigate the relationships between surface pressure distributions on selected fruits and their respective internal textural qualities. A piezoelectric dot-matrix pressure transducer was developed for the purpose of acquiring time-sampled pressure profiles during impact. The acquired data was transferred into a personal computer and accurate visualization of animated data were presented. Preliminary test with 10 apples has been performed. Measurement were made by the contact-pressure transducer in two different positions. Complementary measurements were made on the same apples by using the Firmalon and Magness Taylor (MT) testers. Three-layer neural network was designed. 2/3 of the contact-pressure data were used as training input data and corresponding MT data as training target data. The remaining data were used as NN checking data. Six samples randomly chosen from the ten measured samples and their corresponding Firmalon values were used as the NN training and target data, respectively. The remaining four samples' data were input to the NN. The NN results consistent with the Firmness Tester values. So, if more training data would be obtained, the output should be more accurate. In addition, the Firmness Tester values do not consistent with MT firmness tester values. The NN method developed in this study appears to be a useful tool to emulate the MT Firmness test results without destroying the apple samples. To get more accurate estimation of MT firmness a much larger training data set is required. When the larger sensitive area of the pressure sensor being developed in this project becomes available, the entire contact 'shape' will provide additional information and the neural network results would be more accurate. It has been shown that the impact information can be utilized in the determination of internal quality factors of fruit. Until now,