An algorithm for fast convergence in training neural networks

Wilamowski, Bogdan M.; İplikçi, Serdar; Kaynak, Okyay; Efe, Mehmet Önder

doi:10.1109/ijcnn.2001.938431

Cited by 124 publications

(92 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Bayes' rule automatically regulates the network's performance for smoother running to avoid any overfitting. Further discussion on LM algorithm and Bayesian regularisation is beyond the scope of this article, the interested reader may find detailed description in the literature [17][18][19][20].…”

Section: Predicting Yarn Evenness By Annmentioning

confidence: 99%

Prediction of Blended Yarn Evenness and Tensile Properties by Using Artificial Neural Network and Multiple Linear Regression

Malik¹,

Farooq

Gereke

et al. 2016

Autex Research Journal

View full text Add to dashboard Cite

The present research work was carried out to develop the prediction models for blended ring spun yarn evenness and tensile parameters using artificial neural networks (ANNs) and multiple linear regression (MLR). Polyester/cotton blend ratio, twist multiplier, back roller hardness and break draft ratio were used as input parameters to predict yarn evenness in terms of CVm% and yarn tensile properties in terms of tenacity and elongation. Feed forward neural networks with Bayesian regularisation support were successfully trained and tested using the available experimental data. The coefficients of determination of ANN and regression models indicate that there is a strong correlation between the measured and predicted yarn characteristics with an acceptable mean absolute error values. The comparative analysis of two modelling techniques shows that the ANNs perform better than the MLR models. The relative importance of input variables was determined using rank analysis through input saliency test on optimised ANN models and standardised coefficients of regression models. These models are suitable for yarn manufacturers and can be used within the investigated knowledge domain.

show abstract

Section: Predicting Yarn Evenness By Annmentioning

confidence: 99%

Prediction of Blended Yarn Evenness and Tensile Properties by Using Artificial Neural Network and Multiple Linear Regression

Malik¹,

Farooq

Gereke

et al. 2016

Autex Research Journal

View full text Add to dashboard Cite

show abstract

“…Artificial neural networks have large numbers of computational units called neurons, connected in a massively parallel structure and do not need an explicit formulation of the mathematical or physical relationships of the handled problem [5,6,[8][9][10][11]. The most commonly used ANNs are the feed-forward neural networks [11], which are designed with one input layer, one output layer and hidden layers [8][9][10].…”

Section: Artificial Neural Networkmentioning

confidence: 99%

“…The optimization method chosen in this work was the Levenberg-Marquart algorithm [5][6][7], as mentioned earlier. The testing set is used during the adjustment of the network's synaptic weights to evaluate the algorithms performance on the data not used for tuning and stop the tuning if the error on the testing set increases.…”

Section: Artificial Neural Networkmentioning

confidence: 99%

“…The disadvantage of FNNs is the determination of the ideal number of neurons in the hidden layer(s); few neurons produce a network with low precision and a higher number leads to overfitting and bad quality of interpolation and extrapolation. The use of techniques such as Bayesian regularization, together with a Levenberg-Marquardt algorithm, can help overcome this problem [5][6][7].…”

Section: Artificial Neural Networkmentioning

confidence: 99%

“…Among various ANNs reported in the literature, the feed-forward (back propagation) neural network (FNN) method with a modified LevenbergMarquardt algorithm [5][6][7] is used, which can provide fast and accurate estimation of hydrate stability zones. The method has been developed using experimental data from the literature [1].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Predicting the Hydrate Stability Zones of Natural Gases Using Artificial Neural Networks

Chapoy

Mohammadi

Richon

2007

Oil & Gas Science and Technology - Rev. IFP

207

View full text Add to dashboard Cite

Résumé -Prévision des zones de stabilité d'hydrates de gaz naturels par utilisation de réseaux de neurones artificiels -Une méthode par réseaux de neurones artificiels avec alimentation dans le sens direct faisant appel à 19 variables d'entrée (température, structure de l'hydrate, composition du gaz et concentration de l'inhibiteur en phase aqueuse) et 35 neurones dans une phase cachée unique, a été développée pour estimer les pressions de dissociation d'hydrates de gaz naturels en présence ou non d'inhibiteur en solution aqueuse. Le modèle a été développé en faisant appel à 3296 données de dissociation d'hydrates collectées dans la littérature. La fiabilité de la méthode a été vérifiée par utilisation de données expérimentales indépendantes (non utilisées pour l'apprentissage du réseau de neurones et le développement du modèle). On montre que les résultats de calculs (prédictions) sont en accord acceptable avec les valeurs expérimentales prouvant ainsi la capacité des réseaux de neurones pour l'estimation des zones de stabilité d'hydrates de gaz naturels. Abstract -Predicting the Hydrate Stability Zones of Natural Gases Using Artificial Neural Networks

show abstract

Novel procedure for reliability-based cost optimization of seismically isolated structures for the protection of critical equipment: A case study using single curved surface sliders

Moeindarbari

Taghikhany

2017

Struct Control Health Monit

View full text Add to dashboard Cite

SummaryThe optimized reliability-based design of seismic isolated structures requires a consideration of the failure probability of the system, in addition to the optimization objectives. The stochastic nature of a given system (e.g., the probable input ground motion) must therefore be properly included in the analysis model. To address this challenge, a novel procedure is developed and examined on a 3-story isolated concrete building model, to minimize the total construction cost of the system with regards to protecting sensitive equipment located within the structure.In this procedure, the structure performance and reliability were first evaluated using time-consuming Monte Carlo simulations. We then employed artificial neural networks as a response surface to facilitate the prediction of the failure probability for the supposed structure. To simulate seismic excitations, we generated artificial ground motion combining random high-frequency and long-period components.Also, a newly developed sensitivity analysis method was used to identify the critical uncertain parameters of the system. Finally, by using a simulated annealing algorithm, we determined the optimal design variables of the structure and isolation system for a range of desired probabilities of failure.The optimal results indicate that, for different target failure probability ranges, some design variables are more significant than others. KEYWORDS base isolation, critical equipment protection, neural network, reliability based cost optimization, sensitivity analysis, spectral non-stationarity

show abstract

An algorithm for fast convergence in training neural networks

Cited by 124 publications

References 13 publications

Prediction of Blended Yarn Evenness and Tensile Properties by Using Artificial Neural Network and Multiple Linear Regression

Prediction of Blended Yarn Evenness and Tensile Properties by Using Artificial Neural Network and Multiple Linear Regression

Predicting the Hydrate Stability Zones of Natural Gases Using Artificial Neural Networks

Novel procedure for reliability-based cost optimization of seismically isolated structures for the protection of critical equipment: A case study using single curved surface sliders

Contact Info

Product

Resources

About