scholarly journals Forecast of consumer behaviour based on neural networks models comparison

2012 ◽  
Vol 60 (2) ◽  
pp. 437-442 ◽  
Author(s):  
Michael Štencl ◽  
Ondřej Popelka ◽  
Jiří Šťastný
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Performance Metrics ◽  
Network Models ◽  
Household Consumption ◽  
Feed Forward Neural Network ◽  
Neural Network Models ◽  
Common Error ◽  
Artificial Neural

The aim of this article is comparison of accuracy level of forecasted values of several artificial neural network models. The comparison is performed on datasets of Czech household consumption values. Several statistical models often resolve this task with more or fewer restrictions. In previous work where models’ input conditions were not so strict and model with missing data was used (the time series didn’t contain many values) we have obtained comparably good results with artificial neural networks. Two views – practical and theoretical, motivate the purpose of this study. Forecasting models for medium term prognosis of the main trends of Czech household consumption is part of the faculty research design grant MSM 6215648904/03/02 (Sub-task 5.3) which defines the practical purpose. Testing of nonlinear autoregressive artificial neural network model compared with feed-forward neural network and radial basis function neural network defines the theoretical purpose. The performance metrics of the models were evaluated using a combination of common error metrics, namely Correlation Coefficient and Mean Square Error, together with the number of epochs and/or main prediction error.

Artificial Neural Networks

2011 ◽  
pp. 222-243
Author(s):  
Joarder Kamruzzaman ◽  
Ruhul Sarker
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Artificial Neural Networks ◽  
Network Models ◽  
Neural Network Models ◽  
Network Applications ◽  
The Neural Network ◽  
Artificial Neural ◽  
Neural Network Applications

The primary aim of this chapter is to present an overview of the artificial neural network basics and operation, architectures, and the major algorithms used for training the neural network models. As can be seen in subsequent chapters, neural networks have made many useful contributions to solve theoretical and practical problems in finance and manufacturing areas. The secondary aim here is therefore to provide a brief review of artificial neural network applications in finance and manufacturing areas.

Comparing Conventional and Artificial Neural Network Models for the Pricing of Options

2002 ◽  
pp. 220-235 ◽  
Author(s):  
Paul Lajbcygier
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Artificial Neural Networks ◽  
Input Parameter ◽  
Network Models ◽  
Neural Network Models ◽  
Artificial Neural ◽  
Conventional Models ◽  
Artificial Neural Network Models

The pricing of options on futures is compared using conventional models and artificial neural networks. This work demonstrates superior pricing accuracy using the artificial neural networks in an important subset of the input parameter set.

Artificial Neural Networks

2006 ◽  
pp. 1-27 ◽  
Author(s):  
Joarder Kamruzzaman ◽  
Ruhul A. Sarker
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Artificial Neural Networks ◽  
Network Models ◽  
Neural Network Models ◽  
Network Applications ◽  
The Neural Network ◽  
Artificial Neural ◽  
Neural Network Applications

The primary aim of this chapter is to present an overview of the artificial neural network basics and operation, architectures, and the major algorithms used for training the neural network models. As can be seen in subsequent chapters, neural networks have made many useful contributions to solve theoretical and practical problems in finance and manufacturing areas. The secondary aim here is therefore to provide a brief review of artificial neural network applications in finance and manufacturing areas.

scholarly journals Accurate Prediction of Concentration Changes in Ozone as an Air Pollutant by Multiple Linear Regression and Artificial Neural Networks

Mathematics ◽  
2021 ◽  
Vol 9 (4) ◽  
pp. 356 ◽  
Author(s):  
Svajone Bekesiene ◽  
Ieva Meidute-Kavaliauskiene ◽  
Vaida Vasiliauskiene
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Linear Regression ◽  
Network Models ◽  
Air Pollutant ◽  
Neural Network Models ◽  
Ozone Concentrations ◽  
Concentration Changes ◽  
Artificial Neural

This study considers the usage of multilinear regression and artificial neural network modelling to forecast ozone concentrations with regard to weather-related indicators (wind speed, wind direction, relative humidity and temperature). Initial data were obtained by measuring the meteorological parameters using the PC Radio Weather Station. Ozone concentrations near high-voltage lines were measured using RS1003 and at a 220 m distance using ML9811. Neural network models such as the multilayer perceptron and radial basis function neural networks were constructed. The prognostic capacities of the designed models were assessed by comparing the result data by way of the square of the coefficient of multiple correlations (R2) and mean square error (MSE) values. The number of hidden neurons was optimised by decreasing an error function that recorded the number of units in the hidden layers to the precision of the expanded networks. The neural software IBM SPSS 26v was used for artificial neural network (ANN) modelling. The study demonstrated that the linear regression modelling approach was lacking in its capacity to predict the investigated ozone concentrations by used parameters, whereas the use of an ANN offered more precise outcomes. The conducted tests’ results established the strength of the designed artificial neural network models with irrelevant differences between detected and forecasted data.

scholarly journals Automatic migraine classification using artificial neural networks

F1000Research ◽  
2020 ◽  
Vol 9 ◽  
pp. 618
Author(s):  
Paola A. Sanchez-Sanchez ◽  
José Rafael García-González ◽  
Juan Manuel Rúa Ascar
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Artificial Neural Networks ◽  
Network Models ◽  
Neural Network Models ◽  
Accuracy And Precision ◽  
Artificial Neural ◽  
Artificial Neural Network Models ◽  
Selection Of

Background: Previous studies of migraine classification have focused on the analysis of brain waves, leading to the development of complex tests that are not accessible to the majority of the population. In the early stages of this pathology, patients tend to go to the emergency services or outpatient department, where timely identification largely depends on the expertise of the physician and continuous monitoring of the patient. However, owing to the lack of time to make a proper diagnosis or the inexperience of the physician, migraines are often misdiagnosed either because they are wrongly classified or because the disease severity is underestimated or disparaged. Both cases can lead to inappropriate, unnecessary, or imprecise therapies, which can result in damage to patients’ health. Methods: This study focuses on designing and testing an early classification system capable of distinguishing between seven types of migraines based on the patient’s symptoms. The methodology proposed comprises four steps: data collection based on symptoms and diagnosis by the treating physician, selection of the most relevant variables, use of artificial neural network models for automatic classification, and selection of the best model based on the accuracy and precision of the diagnosis. Results: The artificial neural network models used provide an excellent classification performance, with accuracy and precision levels >97% and which exceed the classifications made using other model, such as logistic regression, support vector machines, nearest neighbor, and decision trees. Conclusions: The implementation of migraine classification through artificial neural networks is a powerful tool that reduces the time to obtain accurate, reliable, and timely clinical diagnoses.

scholarly journals PRONÓSTICO DE CAUDALES MEDIOS MENSUALES DEL RIO ILAVE USANDO MODELOS DE REDES NEURONALES ARTIFICIALES

2014 ◽  
Vol 16 (01) ◽  
Author(s):  
EFRAIN LUJANO LAURA ◽  
APOLINARIO LUJANO ◽  
JOSÉ PITÁGORAS QUISPE ◽  
RENÉ LUJANO
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Back Propagation ◽  
Network Models ◽  
Mean Square ◽  
Neural Network Models ◽  
Artificial Neural ◽  
Del Rio ◽  
Artificial Neural Network Models

<h4 class="text-primary">Resumen</h4><p style="text-align: justify;">La presente investigación se realizó en la cuenca del río Ilave, ubicado dentro de la región Hidrográfica del Titicaca (Perú), teniendo como objetivo pronosticar los caudales medios mensuales del rio Ilave usando Modelos de Redes Neuronales Artificiales, aplicado al problema del pronóstico mensual de esta variable, cuyo resultado puede emplearse en la planificación y gestión de los recursos hídricos en cuencas hidrográficas. La información hidrometeorológica utilizada, corresponde al Servicio Nacional de Meteorología e Hidrología con un periodo de registro de 1965 al 2007, de donde se plantearon 06 modelos que están en función de precipitaciones y caudales, cuya fase de entrenamiento, validación y prueba, se realizaron con el 70%, 15% y 15% del total de datos respectivamente, con una red de entrenamiento designada Perceptrón Multicapa (MLP) y el algoritmo «back-propagatión». La significación estadística de los indicadores de desempeño de eficiencia de Nash-Sutcliffe (NSE) y la raíz del error cuadrático medio (RMSE), fueron evaluados usando el método de bootstrap incorporado en el código FITEVAL y como indicadores complementarios de evaluación tradicional, el coeficiente de determinación (R2) y el error cuadrático medio normalizado (ECMN). Los resultados de validación y prueba indican calificativos de buenos a muy buenos, así tenemos que en la fase de pronóstico para los modelos seleccionados MRNA5, MRNA2 y MRNA3, los coeficientes de Eficiencia de Nash-Sutcliffe son de 88.0%, 87.9% y 87.1%; la raíz del error medio cuadrático son de 18.87%, 18.96% y 19.56% respectivamente. Se concluye que el pronóstico de caudales medios mensuales del río Ilave utilizando modelos de Redes Neuronales Artificiales, muestran un buen desempeño en la estimación de fenómenos de comportamiento no lineal como los caudales.</p><p><strong>PALABRAS CLAVE: </strong>* Backpropagation * caudales medios * redes neuronales artificiales río * Ilave</p><h4 class="text-primary">ABSTRACT</h4><p><strong>AVERAGE FLOW-MONTHLY FORECAST OF THE ILAVE RIVER USING ARTIFICIAL NEURAL NETWORK MODELS</strong></p><p style="text-align: justify;">This research was conducted in the Ilave river basin located within the hydrographic region of Titicaca (Peru), aiming to predict the average monthly flow of the river Ilave usingArtificial Neural Networks models applied to forecast the monthly variable flow of this river. The results of this type of forecasting can be used in the planning and management of water resources in river basins. The hydrometeorological information used, corresponds to the National Meteorological and Hydrological Service registries between 1965 – 2007. 06 models were proposed that are based on rainfall and river flow, whose training, validation and testing phases were realized with 70%, 15% and 15% of the total data respectively. A training network titled Multilayer Perception (MLP) as well as algorithm and «back -propagation»techniques were used. The statistical significance of the performance indicators Nash (NSE) and the Root Mean Square Error (RMSE), were assessed using the bootstrap method incorporated in the FITEVAL code. The coefficient of determination (R2) and Normalized Root Mean Square Error (NRMSE) were used as complementary to indicators of traditional assessment. The results of test descriptions and validation indicate good to very good results, so in the forecast phase for selected models MRNA5, MRNA2 and MRNA3, Nash coefficients are 88.0%, 87.9% and 87.1%; mean square root error are 18.87%, 18.96% and 19.56% respectively. We conclude that the average monthly flow forecast of the river Ilave, using Artificial Neural Network models, show a good performance in estimating nonlinear phenomena such as flow behavior.</p><p><strong>KEY WORDS: </strong>* artificial neural networks * back propagation * Ilave river * mean flows</p>

scholarly journals Photovoltaic Panel Efficiency Estimation with Artificial Neural Networks: Samples of Adiyaman, Malatya, and Sanliurfa

2019 ◽  
Vol 2019 ◽  
pp. 1-12
Author(s):  
Yasin Icel ◽  
Mehmet Salih Mamis ◽  
Abdulcelil Bugutekin ◽  
Mehmet Ismail Gursoy
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Artificial Neural Networks ◽  
Network Models ◽  
Estimation Accuracy ◽  
Neural Network Models ◽  
Photovoltaic Panels ◽  
Artificial Neural ◽  
Artificial Neural Network Models

The amount of electric energy produced by photovoltaic panels depends on air temperature, humidity rate, wind velocity, photovoltaic module temperature, and particularly solar radiation. Being aware of the behaviour patterns of the panels to be used in project and planning works regarding photovoltaic applications will set forth a realistic expense form; therefore, erroneous investments will be avoided, and the country budget will benefit from added value. The power ratings obtained from the photovoltaic panels and the environmental factors were measured and recorded for a year by the measurement stations established in three diverse regions (Adiyaman-Malatya-Sanliurfa). In the developed artificial neural network models, the estimation accuracy was 99.94%. Furthermore, by taking the data of the General Directorate of Meteorology as a reference, models of artificial neural networks were developed using the data from Adiyaman province for training; by using Malatya and Sanliurfa as test data, 99.57% estimation accuracy was achieved. With the artificial neural network models developed as a result of the study, the energy efficiency for the photovoltaic energy systems desired to be established by using meteorological parameters such as temperature, humidity, wind, and solar radiation of various regions anywhere in the world can be estimated with high accuracy.

scholarly journals Prediction of Stress in Power Transformer Winding Conductors Using Artificial Neural Networks: Hyperparameter Analysis

Energies ◽  
2021 ◽  
Vol 14 (14) ◽  
pp. 4242
Author(s):  
Fausto Valencia ◽  
Hugo Arcos ◽  
Franklin Quilumba
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Finite Element ◽  
Power Transformer ◽  
Network Models ◽  
Activation Function ◽  
Neural Network Models ◽  
Transformer Winding ◽  
Artificial Neural ◽  
Femm Software

The purpose of this research is the evaluation of artificial neural network models in the prediction of stresses in a 400 MVA power transformer winding conductor caused by the circulation of fault currents. The models were compared considering the training, validation, and test data errors’ behavior. Different combinations of hyperparameters were analyzed based on the variation of architectures, optimizers, and activation functions. The data for the process was created from finite element simulations performed in the FEMM software. The design of the Artificial Neural Network was performed using the Keras framework. As a result, a model with one hidden layer was the best suited architecture for the problem at hand, with the optimizer Adam and the activation function ReLU. The final Artificial Neural Network model predictions were compared with the Finite Element Method results, showing good agreement but with a much shorter solution time.

Sign in / Sign up

Export Citation Format

Share Document