875 resultados para Machine Learning Robotics Artificial Intelligence Bayesian Networks


Relevância:

100.00% 100.00%

Publicador:

Resumo:

Cuttings return analysis is an important tool to detect and prevent problems during the petroleum well drilling process. Several measurements and tools have been developed for drilling problems detection, including mud logging, PWD and downhole torque information. Cuttings flow meters were developed in the past to provide information regarding cuttings return at the shale shakers. Their use, however, significantly impact the operation including rig space issues, interferences in geological analysis besides, additional personel required. This article proposes a non intrusive system to analyze the cuttings concentration at the shale shakers, which can indicate problems during drilling process, such as landslide, the collapse of the well borehole walls. Cuttings images are acquired by a high definition camera installed above the shakers and sent to a computer coupled with a data analysis system which aims the quantification and closure of a cuttings material balance in the well surface system domain. No additional people at the rigsite are required to operate the system. Modern Artificial intelligence techniques are used for pattern recognition and data analysis. Techniques include the Optimum-Path Forest (OPF), Artificial Neural Network using Multilayer Perceptrons (ANN-MLP), Support Vector Machines (SVM) and a Bayesian Classifier (BC). Field test results conducted on offshore floating vessels are presented. Results show the robustness of the proposed system, which can be also integrated with other data to improve the efficiency of drilling problems detection. Copyright 2010, IADC/SPE Drilling Conference and Exhibition.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

The applications of Automatic Vowel Recognition (AVR), which is a sub-part of fundamental importance in most of the speech processing systems, vary from automatic interpretation of spoken language to biometrics. State-of-the-art systems for AVR are based on traditional machine learning models such as Artificial Neural Networks (ANNs) and Support Vector Machines (SVMs), however, such classifiers can not deal with efficiency and effectiveness at the same time, existing a gap to be explored when real-time processing is required. In this work, we present an algorithm for AVR based on the Optimum-Path Forest (OPF), which is an emergent pattern recognition technique recently introduced in literature. Adopting a supervised training procedure and using speech tags from two public datasets, we observed that OPF has outperformed ANNs, SVMs, plus other classifiers, in terms of training time and accuracy. ©2010 IEEE.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

This paper introduces a methodology for predicting the surface roughness of advanced ceramics using Adaptive Neuro-Fuzzy Inference System (ANFIS). To this end, a grinding machine was used, equipped with an acoustic emission sensor and a power transducer connected to the electric motor rotating the diamond grinding wheel. The alumina workpieces used in this work were pressed and sintered into rectangular bars. Acoustic emission and cutting power signals were collected during the tests and digitally processed to calculate the mean, standard deviation, and two other statistical data. These statistics, as well the root mean square of the acoustic emission and cutting power signals were used as input data for ANFIS. The output values of surface roughness (measured during the tests) were implemented for training and validation of the model. The results indicated that an ANFIS network is an excellent tool when applied to predict the surface roughness of ceramic workpieces in the grinding process.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Musical genre classification has been paramount in the last years, mainly in large multimedia datasets, in which new songs and genres can be added at every moment by anyone. In this context, we have seen the growing of musical recommendation systems, which can improve the benefits for several applications, such as social networks and collective musical libraries. In this work, we have introduced a recent machine learning technique named Optimum-Path Forest (OPF) for musical genre classification, which has been demonstrated to be similar to the state-of-the-art pattern recognition techniques, but much faster for some applications. Experiments in two public datasets were conducted against Support Vector Machines and a Bayesian classifier to show the validity of our work. In addition, we have executed an experiment using very recent hybrid feature selection techniques based on OPF to speed up feature extraction process. © 2011 International Society for Music Information Retrieval.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Voice-based user interfaces have been actively pursued aiming to help individuals with motor impairments, providing natural interfaces to communicate with machines. In this work, we have introduced a recent machine learning technique named Optimum-Path Forest (OPF) for voice-based robot interface, which has been demonstrated to be similar to the state-of-the-art pattern recognition techniques, but much faster. Experiments were conducted against Support Vector Machines, Neural Networks and a Bayesian classifier to show the OPF robustness. The proposed architecture provides high accuracy rates allied with low computational times. © 2012 IEEE.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Nowadays, organizations face the problem of keeping their information protected, available and trustworthy. In this context, machine learning techniques have also been extensively applied to this task. Since manual labeling is very expensive, several works attempt to handle intrusion detection with traditional clustering algorithms. In this paper, we introduce a new pattern recognition technique called Optimum-Path Forest (OPF) clustering to this task. Experiments on three public datasets have showed that OPF classifier may be a suitable tool to detect intrusions on computer networks, since it outperformed some state-of-the-art unsupervised techniques. © 2012 IEEE.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Grinding is a parts finishing process for advanced products and surfaces. However, continuous friction between the workpiece and the grinding wheel causes the latter to lose its sharpness, thus impairing the grinding results. This is when the dressing process is required, which consists of sharpening the worn grains of the grinding wheel. The dressing conditions strongly affect the performance of the grinding operation; hence, monitoring them throughout the process can increase its efficiency. The objective of this study was to estimate the wear of a single-point dresser using intelligent systems whose inputs were obtained by the digital processing of acoustic emission signals. Two intelligent systems, the multilayer perceptron and the Kohonen neural network, were compared in terms of their classifying ability. The harmonic content of the acoustic emission signal was found to be influenced by the condition of dresser, and when used to feed the neural networks it is possible to classify the condition of the tool under study.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Some machine learning methods do not exploit contextual information in the process of discovering, describing and recognizing patterns. However, spatial/temporal neighboring samples are likely to have same behavior. Here, we propose an approach which unifies a supervised learning algorithm - namely Optimum-Path Forest - together with a Markov Random Field in order to build a prior model holding a spatial smoothness assumption, which takes into account the contextual information for classification purposes. We show its robustness for brain tissue classification over some images of the well-known dataset IBSR. © 2013 Springer-Verlag.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Pós-graduação em Agronomia (Energia na Agricultura) - FCA

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Pós-graduação em Engenharia Elétrica - FEIS

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Pós-graduação em Ciências Biológicas (Genética) - IBB

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Pós-graduação em Engenharia Mecânica - FEG

Relevância:

100.00% 100.00%

Publicador:

Resumo:

A adoção de sistemas digitais de radiodifusão sonora, que estão em fase de testes no país, permite realizar novos estudos visando um melhor planejamento para a implementação dessas novas emissoras. O que significa reavaliar os principais modelos de radiopropagação existentes ou propor novas alternativas para atender as demandas inerentes dos sistemas digitais. Os modelos atuais, conforme Recomendações ITU-R P. 1546 e ITU-R P. 1812, não condizem fielmente com a realidade de algumas regiões do Brasil, principalmente com as regiões de clima tropical, como a Região Amazônica, seja pelo elevado índice pluviométrico seja pela vasta flora existente. A partir dos modelos adequados ao canal de propagação, torna-se viável desenvolver ferramentas de planejamento de cobertura mais precisas e eficientes. A utilização destas ferramentas é cabível tanto para a ANATEL, para a elaboração dos planos básicos de distribuição de canais quanto para os radiodifusores. No presente trabalho é apresentada uma metodologia utilizando a inteligência computacional, baseada em Inferênciass Baysianas, para predição da intensidade de campo elétrico, a qual pode ser aplicada ao planejamento ou expansão de áreas de cobertura em sistemas de radiodifusão para frequências na faixa de ondas médias (de 300 kHz a 3MHz). Esta metodologia gera valores de campo elétrico estimados a partir dos valores de altitude do terreno (através de análises de tabelas de probabilidade condicional) e estabelece a comparação destes com valores de campo elétrico medidos. Os dados utilizados neste trabalho foram coletados na região central do Brasil, próximo à cidade de Brasília. O sinal transmitido era um sinal de rádio AM transmitido na frequência de 980 kHz. De posse dos dados coletados durante as campanhas de medição, foram realizadas simulações utilizando tabelas de probabilidade condicional geradas por Inferências Bayesianas. Assim, é proposto um método para predizer valores de campo elétrico com base na correlação entre o campo elétrico medido e altitude, através da utilização de inteligência computacional. Se comparados a inúmeros trabalhos existentes na literatura que têm o mesmo objetivo, os resultados encontrados neste trabalho validam o uso da metodologia para determinar o campo elétrico de radiodifusão sonora em ondas médias utilizando Inferências Bayesianas.