scholarly journals Tracking Turbulent Coherent Structures by Means of Neural Networks

Energies ◽  
2021 ◽  
Vol 14 (4) ◽  
pp. 984
Author(s):  
Jose J. Aguilar-Fuertes ◽  
Francisco Noguero-Rodríguez ◽  
José C. Jaen Ruiz ◽  
Luis M. García-RAffi ◽  
Sergio Hoyas
Keyword(s):  
Turbulent Flows ◽  
Short Term Memory ◽  
Computational Cost ◽  
Multi Layer Perceptron ◽  
Short Term ◽  
Turbulent Coherent Structures ◽  
Geometrical Features ◽  
Consecutive Time ◽  
Long Short Term Memory ◽  
Logarithmic Layer

The behaviours of individual flow structures have become a relevant matter of study in turbulent flows as the computational power to allow their study feasible has become available. Especially, high instantaneous Reynolds Stress events have been found to dominate the behaviour of the logarithmic layer. In this work, we present a viability study where two machine learning solutions are proposed to reduce the computational cost of tracking such structures in large domains. The first one is a Multi-Layer Perceptron. The second one uses Long Short-Term Memory (LSTM). Both of the methods are developed with the objective of taking the the structures’ geometrical features as inputs from which to predict the structures’ geometrical features in future time steps. Some of the tested Multi-Layer Perceptron architectures proved to perform better and achieve higher accuracy than the LSTM architectures tested, providing lower errors on the predictions and achieving higher accuracy in relating the structures in the consecutive time steps.

scholarly journals Inferência dos Perfis de Infusão em Sistemas Intravenosos: Uma Abordagem Empregando Técnicas de Aprendizagem de Máquina

2021 ◽  
Author(s):  
Fabrício Ferreira ◽  
Felipe Gruendemann ◽  
Ricardo Araujo ◽  
Adenauer Yamin ◽  
Luciano Agostini
Keyword(s):  
Short Term Memory ◽  
Multi Layer Perceptron ◽  
Short Term ◽  
Term Memory ◽  
Long Short Term Memory

Os procedimentos de infusão intravenosa estão entre os mais usuais em hospitais e têm potencial para gerar alta ocorrência de eventos adversos. No entanto, as infusões intravenosas ainda não têm a sua verificação automatizada. Considerando este cenário, este trabalho propõe uma nova abordagem para reduzir eventos adversos em procedimentos intravenosos utilizando Aprendizado de Máquina para permitir uma inferência autônoma e registro dos perfis de infusões intravenosas. Dois regressores baseados em redes neurais foram avaliados: Multi-Layer Perceptron e Long-Short Term Memory. A avaliação dos modelos regressão, para as inferências dos perfis de administração de medicamentos intravenosos, obtiveram resultados promissores.

Kajian Penggunaan Data Log Mahasiswa untuk Berbagai Permasalahan Analisis Pembelajaran

2020 ◽  
Vol 9 (4) ◽  
pp. 365-374
Author(s):  
Sri Suning Kusumawardani ◽  
Syukron Abu Ishaq Alfarozi
Keyword(s):  
Support Vector Machine ◽  
Learning Analytics ◽  
Short Term Memory ◽  
Drop Out ◽  
Support Vector ◽  
Multi Layer Perceptron ◽  
Short Term ◽  
Term Memory ◽  
Long Short Term Memory ◽  
Data Log

Pada saat ini, penyelenggaraan sistem pembelajaran daring menjadi hal yang penting di tengah pandemi untuk menekan persebaran virus COVID-19. Namun, sistem ini sangat sulit menjaga motivasi dan tingkat keterlibatan mahasiswa karena tidak ada interaksi langsung antara pengajar dengan mahasiswa. Makalah ini meninjau penggunaan data log mahasiswa untuk kebutuhan analisis pembelajaran guna memprediksi kinerja atau kecenderungan drop-out mahasiswa dari suatu mata kuliah dengan melihat pada data log interaksi mahasiswa dengan sistem dan data demografis mahasiswa menggunakan suatu data terbuka, yaitu Open University Learning Analytics Dataset (OULAD). Dari tinjauan beberapa artikel penelitian yang merujuk pada dataset tersebut, ada beberapa hal yang perlu ditinjau: 1) permasalahan yang sering diangkat, yaitu prediksi kecenderungan gagal dari mata kuliah tertentu, prediksi kinerja, dan prediksi keterlibatan mahasiswa; 2) fitur yang digunakan pada saat pemodelan, yaitu fitur demografis dan interaksi, baik yang diringkas secara harian atau mingguan dengan berbagai representasi fitur; 3) metode analisis pembelajaran yang secara khusus menggunakan metode pembelajaran mesin yang sering digunakan, yaitu Decision Tree (DT), Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), dan Long Short-Term Memory (LSTM). Makalah ini juga mendiskusikan proses mitigasi dari mahasiswa yang berisiko, perancangan sistem data yang mendukung analisis pembelajaran, dan permasalahan yang sering ditemui pada saat proses pemodelan.

scholarly journals Aplicação de redes neurais artificiais para previsão de enchentes no rio Itajaí-Açu em Blumenau, SC, Brasil

2021 ◽  
Vol 12 (4) ◽  
pp. 686-696
Author(s):  
Gabriel Barreto Alberton ◽  
Dirceu Luis Severo ◽  
Mateus Nascimento Vieira de Melo ◽  
Hélio Potelicki ◽  
Andreza Sartori
Keyword(s):  
Short Term Memory ◽  
Multi Layer Perceptron ◽  
Short Term ◽  
Santa Catarina ◽  
Term Memory ◽  
Long Short Term Memory

As redes neurais artificiais (RNA) têm sido utilizadas com sucesso em previsões de variáveis baseadas em acontecimentos anteriores, porém, escassos são os estudos sobre a aplicação dessa solução à previsão de níveis de rio em eventos de enchentes. Este estudo teve como objetivo avaliar a aplicação de RNA para previsão em curto prazo dos níveis do rio Itajaí-Açu no município de Blumenau, Santa Catarina, Brasil. O município foi escolhido como área de estudo por seu extenso histórico de inundações. Utilizou-se para o treinamento das redes dados de chuva e de nível do rio das estações telemétricas do Sistema Nacional de Informações sobre Recursos Hídricos (SNIRH) da Agência Nacional de Águas (ANA) localizadas na bacia hidrográfica do rio Itajaí-Açu. Ambos dados apresentam frequência de 15 min. Foram selecionados 7 eventos hidrológicos de alerta registrados pela estação limnimétrica instalada no município de Blumenau. Os dados foram coletados e reunidos de acordo com sua localização e tipo, e foram escalados à mesma unidade de medida: cm para níveis e mm para dados de precipitação. Foram utilizados dois tipos de redes: Long Short Term Memory (LSTM) e Multi Layer Perceptron (MLP). Para avaliação de desempenho dos modelos, utilizou-se os seguintes parâmetros: coeficiente de determinação (R2), o Coeficiente de Eficiência de Nash-Sutcliffe (NSE), a Raíz do Erro Quadrático Médio (RMSE), Erro Quadrático Médio (MSE), o Erro Médio Absoluto (MAE) e a Média Percentual Absoluta do Erro (MAPE). Para o modelo com melhor desempenho – modelo LSTM com horizonte de previsão de 6 h – obteve-se: R2 = 0,996594; NSE = 0,9995548; RMSE = 7,72 cm; MSE = 59,65 cm; MAE = 4,82 cm; MAPE = 1,89 %; e MSE val. = 0,000035. O estudo evidenciou que o modelo LSTM, com simples pré-processamento, é capaz de prever o nível do Itajaí-Açu durante eventos extremos de cheia, com alta precisão, apresentando resultados melhores em comparação com o modelo MLP. Este estudo apresenta uma proposta de solução de modelo de previsão de níveis viável, passível de aplicação como ferramenta de previsão em tempo real para a área de estudo. Este trabalho contribui para o desenvolvimento de sistemas de apoio à gestão de recursos hídricos e para mitigação dos impactos provocados por desastres, abrangendo os âmbitos social, econômico e ambiental.

scholarly journals Real-Time PPG Signal Conditioning with Long Short-Term Memory (LSTM) Network for Wearable Devices

Sensors ◽  
2021 ◽  
Vol 22 (1) ◽  
pp. 164
Author(s):  
Marek Wójcikowski
Keyword(s):  
Heart Rate ◽  
Real Time ◽  
Time Domain ◽  
Short Term Memory ◽  
Computational Cost ◽  
Computational Effort ◽  
Short Term ◽  
Term Memory ◽  
Long Short Term Memory ◽  
Lstm Network

This paper presents an algorithm for real-time detection of the heart rate measured on a person’s wrist using a wearable device with a photoplethysmographic (PPG) sensor and accelerometer. The proposed algorithm consists of an appropriately trained LSTM network and the Time-Domain Heart Rate (TDHR) algorithm for peak detection in the PPG waveform. The Long Short-Term Memory (LSTM) network uses the signals from the accelerometer to improve the shape of the PPG input signal in a time domain that is distorted by body movements. Multiple variants of the LSTM network have been evaluated, including taking their complexity and computational cost into consideration. Adding the LSTM network caused additional computational effort, but the performance results of the whole algorithm are much better, outperforming the other algorithms from the literature.

scholarly journals Forecasting adverse surgical events using self-supervised transfer learning for physiological signals

2021 ◽  
Vol 4 (1) ◽  
Author(s):  
Hugh Chen ◽  
Scott M. Lundberg ◽  
Gabriel Erion ◽  
Jerry H. Kim ◽  
Su-In Lee
Keyword(s):  
Time Series ◽  
Transfer Learning ◽  
Short Term Memory ◽  
Computational Cost ◽  
Physiological Signals ◽  
Short Term ◽  
Physiological Signal ◽  
Data Phase ◽  
Unseen Data ◽  
Long Short Term Memory

AbstractHundreds of millions of surgical procedures take place annually across the world, which generate a prevalent type of electronic health record (EHR) data comprising time series physiological signals. Here, we present a transferable embedding method (i.e., a method to transform time series signals into input features for predictive machine learning models) named PHASE (PHysiologicAl Signal Embeddings) that enables us to more accurately forecast adverse surgical outcomes based on physiological signals. We evaluate PHASE on minute-by-minute EHR data of more than 50,000 surgeries from two operating room (OR) datasets and patient stays in an intensive care unit (ICU) dataset. PHASE outperforms other state-of-the-art approaches, such as long-short term memory networks trained on raw data and gradient boosted trees trained on handcrafted features, in predicting six distinct outcomes: hypoxemia, hypocapnia, hypotension, hypertension, phenylephrine, and epinephrine. In a transfer learning setting where we train embedding models in one dataset then embed signals and predict adverse events in unseen data, PHASE achieves significantly higher prediction accuracy at lower computational cost compared to conventional approaches. Finally, given the importance of understanding models in clinical applications we demonstrate that PHASE is explainable and validate our predictive models using local feature attribution methods.

scholarly journals Enlightening and Predicting the Correlation Around Deep Neural Nets and Cognitive Perceptions

2020 ◽  
Vol 14 ◽  
Keyword(s):  
Short Term Memory ◽  
Neural Nets ◽  
Relational Reasoning ◽  
Multi Layer Perceptron ◽  
Short Term ◽  
Physical Systems ◽  
General Ability ◽  
Complex Reasoning ◽  
Long Short Term Memory ◽  
Scientific Validation

Recently, psychologist has experienced drastic development using statistical methods to analyze the interactions of humans. The intention of past decades of psychological studies is to model how individuals learn elements and types. The scientific validation of such studies is often based on straightforward illustrations of artificial stimuli. Recently, in activities such as recognizing items in natural pictures, strong neural networks have reached or exceeded human precision. In this paper, we present Relevance Networks (RNs) as a basic plug-and-play application with Covolutionary Neural Network (CNN) to address issues that are essentially related to reasoning. Thus our proposed network performs visual answering the questions, superhuman performance and text based answering. All of these have been accomplished by complex reasoning on diverse physical systems. Thus, by simply increasing convolutions, (Long Short Term Memory) LSTMs, and (Multi-Layer Perceptron) MLPs with RNs, we can remove the computational burden from network components that are unsuitable for handling relational reasoning, reduce the overall complexity of the network, and gain a general ability to reason about the relationships between entities and their properties.

scholarly journals A Long Short-Term Memory Neural Network for the Low-Cost Prediction of Soot Concentration in a Time-Dependent Flame

Energies ◽  
2021 ◽  
Vol 14 (5) ◽  
pp. 1394
Author(s):  
Mehdi Jadidi ◽  
Luke Di Liddo ◽  
Seth B. Dworkin
Keyword(s):  
Neural Network ◽  
Short Term Memory ◽  
Volume Fraction ◽  
Soot Formation ◽  
Computational Cost ◽  
Computational Time ◽  
High Fidelity ◽  
Short Term ◽  
Term Memory ◽  
Long Short Term Memory

Particulate matter (soot) emissions from combustion processes have damaging health and environmental effects. Numerical techniques with varying levels of accuracy and computational time have been developed to model soot formation in flames. High-fidelity soot models come with a significant computational cost and as a result, accurate soot modelling becomes numerically prohibitive for simulations of industrial combustion devices. In the present study, an accurate and computationally inexpensive soot-estimating tool has been developed using a long short-term memory (LSTM) neural network. The LSTM network is used to estimate the soot volume fraction (fv) in a time-varying, laminar, ethylene/air coflow diffusion flame with 20 Hz periodic fluctuation on the fuel velocity and a 50% amplitude of modulation. The LSTM neural network is trained using data from CFD, where the network inputs are gas properties that are known to impact soot formation (such as temperature) and the network output is fv. The LSTM is shown to give accurate estimations of fv, achieving an average error (relative to CFD) in the peak fv of approximately 30% for the training data and 22% for the test data, all in a computational time that is orders-of-magnitude less than that of high-fidelity CFD modelling. The neural network approach shows great potential to be applied in industrial applications because it can accurately estimate the soot characteristics without the need to solve the soot-related terms and equations.

scholarly journals Passive Shallow Water Automated Target Recognition using Deep Convolutional Bi directional Long Short Term Memory

2021 ◽  
Vol 71 (1) ◽  
pp. 117-123
Author(s):  
Suraj Kamal ◽  
C. Satheesh Chandran ◽  
H.M. Supriya
Keyword(s):  
Short Term Memory ◽  
Temporal Dynamics ◽  
Target Recognition ◽  
Computational Cost ◽  
Short Term ◽  
Term Memory ◽  
Model Search ◽  
Acoustic Target ◽  
Long Short Term Memory ◽  
Passive Acoustic

The extremely challenging nature of passive acoustic surveillance makes it a key area of research in NavalNon-Co-operative Target Recognition especially in Anti-Submarine Warfare systems. In shallow waters, thecomplex acoustics due to the highly varying ambient background noise as well as the multi-modal propagation in the surface-bottom bounded channel makes surveillance even difficult. In this work, an ensemble of Convolutional Neural Networks and Bidirectional Long Short Term Memory stages employing soft attention is used to effectively capture the spectro-temporal dynamics of the target signature. In order to alleviate the overall computational cost associated with the optimal model search in the extensive hyperparameter space, a recursive model elimination scheme, making frugal use of the available resources, is also proposed. Experimental analysis on acoustic target records, collected from the shallows of Arabian Sea, has yielded encouraging results in terms of model accuracy, precision and recall.

Sign in / Sign up

Export Citation Format

Share Document