Filters








2,174 Hits in 3.4 sec

New training approaches for classification based on evolutionary neural networks. Application to product and sigmoidal units

Antonio J. Tallón-Ballesteros
2014 Inteligencia Artificial  
Specifically, three contributions to train feed-forward neural network models based on evolutionary computation for a classification task are described.  ...  Particularly, two kind of neurons such as product and sigmoidal units have been considered in an independent fashion for the hidden layer.  ...  Acknowledgements This work has been partially subsidized by TIN2007-68084-C02-02, TIN2008-06681-C06-03 and TIN2011-28956-C02-02 projects of the Spanish Inter-Ministerial Commission of Science and Technology  ... 
doi:10.4114/intartif.vol17iss54pp30-34 fatcat:a6ibgtsnczbe3hjmq2ag7ny7wi

Evolutionary product-unit neural networks classifiers

F.J. Martínez-Estudillo, C. Hervás-Martínez, P.A. Gutiérrez, A.C. Martínez-Estudillo
2008 Neurocomputing  
This paper proposes a classification method based on a special class of feed-forward neural network, namely product-unit neural networks.  ...  We apply an evolutionary algorithm to determine the basic structure of the product-unit model and to estimate the coefficients of the model.  ...  Acknowledgments This work has been financed in part by TIN 2005-08386-C05-02 projects of the Spanish Inter-Ministerial Commission of Science and Technology (MICYT) and FEDER funds.  ... 
doi:10.1016/j.neucom.2007.11.019 fatcat:xib2itilsvajjdq5yobxuyi5pi

Architecture Optimization Model for the Deep Neural Network

Kingsley Ukaoha, Efosa Igodan
2019 International Journal of Intelligent Computing and Information Sciences  
In this study, a novel approach that combines an evolutionary genetic algorithm and an optimization algorithm and a supervised deep neural network (Deep-NN) using alternative activation functions with  ...  These tasks though determine the success of building and training an effective and accurate model, are yet to be considered on a deep network having three hidden layers with varying optimized parameters  ...  The most popular approaches to ML are Artificial Neural Network (ANN) and Evolutionary Algorithm (EA).  ... 
doi:10.21608/ijicis.2019.96101 fatcat:lxnzf53b5rdlrkzwizfjh4ltnm

Hybrid Artificial Neural Networks: Models, Algorithms and Data [chapter]

P. A. Gutiérrez, C. Hervás-Martínez
2011 Lecture Notes in Computer Science  
Artificial neural networks (ANNs) constitute a class of flexible nonlinear models designed to mimic biological neural systems.  ...  ANNs are one of the three main components of computational intelligence and, as such, they have been often hybridized from different perspectives.  ...  Other one is based on projection functions, such as sigmoidal unit basis functions, which are part of the MultiLayer Perceptron (MLP) [12] , or product units which generate product unit neural networks  ... 
doi:10.1007/978-3-642-21498-1_23 fatcat:l3huh7hohjcqvokqgp55tg67u4

Multinomial logistic regression and product unit neural network models: Application of a new hybrid methodology for solving a classification problem in the livestock sector

Mercedes Torres, Cesar Hervás, Carlos García
2009 Expert systems with applications  
The application of this methodology involves, first of all, training the coefficients and the basis structure of product unit models using techniques based on artificial neural networks and evolutionary  ...  This work presents a new approach for multi-class pattern recognition based on the hybridization of a linear and nonlinear model.  ...  Acknowledgements The authors gratefully acknowledge the financial support provided by the Spanish Inter-Ministerial Commission of Science and Technology (MICYT) (TIN 2008-06681-C), FEDER funds and the  ... 
doi:10.1016/j.eswa.2009.04.070 fatcat:uus26y2gzjen7dfuyppk6w3lpm

Leveraging Product as an Activation Function in Deep Networks [article]

Luke B. Godfrey, Michael S. Gashler
2018 arXiv   pre-print
Product unit neural networks (PUNNs) are powerful representational models with a strong theoretical basis, but have proven to be difficult to train with gradient-based optimizers.  ...  We present windowed product unit neural networks (WPUNNs), a simple method of leveraging product as a nonlinearity in a neural network.  ...  More recent work has focused on hybridizing PUNNs with sigmoid and RBF networks [12] , [13] , [14] , [15] , and developing improved evolutionary algorithms for training PUNNs [16] , [17] , [18]  ... 
arXiv:1810.08578v1 fatcat:rjk55htxnbbxxdqqrxa6gjzw7q

Combined Projection and Kernel Basis Functions for Classification in Evolutionary Neural Networks [chapter]

P. A. Gutiérrez, C. Hervás, M. Carbonero, J. C. Fernández
2007 Advances in Soft Computing  
function (PRBF) neural networks, and sigmoidal-radial basis function (SRBF) neural networks; and these are compared to the corresponding pure models: product unit neural network (PUNN), multilayer perceptron  ...  An evolutionary algorithm is adapted to this model and applied for learning the architecture, weights and node typology.  ...  Acknowledgments This work has been partially subsidized by TIN 2008-06681-C06-03 project of the Spanish Inter-Ministerial Commission of Science and Technology (MICYT), FEDER funds and the P08-TIC-3745  ... 
doi:10.1007/978-3-540-74972-1_13 dblp:series/asc/GutierrezHCF08 fatcat:rolvv77bcvaj7ebrh7dgl22nsu

Combined projection and kernel basis functions for classification in evolutionary neural networks

P.A. Gutiérrez, C. Hervás, M. Carbonero, J.C. Fernández
2009 Neurocomputing  
function (PRBF) neural networks, and sigmoidal-radial basis function (SRBF) neural networks; and these are compared to the corresponding pure models: product unit neural network (PUNN), multilayer perceptron  ...  An evolutionary algorithm is adapted to this model and applied for learning the architecture, weights and node typology.  ...  Acknowledgments This work has been partially subsidized by TIN 2008-06681-C06-03 project of the Spanish Inter-Ministerial Commission of Science and Technology (MICYT), FEDER funds and the P08-TIC-3745  ... 
doi:10.1016/j.neucom.2008.09.020 fatcat:ootnrc22pvgabm2pbuqnuvwtfy

Evolutionary training of hardware realizable multilayer perceptrons

V. P. Plagianakos, G. D. Magoulas, M. N. Vrahatis
2005 Neural computing & applications (Print)  
The proposed evolutionary strategy does not need gradient related information, it is applicable to a situation where threshold activations are used from the beginning of the training, as in "on-chip" training  ...  , and is able to train networks with integer weights.  ...  Acknowledgements The authors would like to thank the European Social Fund, Operational Program for Educational and Vocational Training II (EPEAEK II), and particularly the Program PY-THAGORAS for funding  ... 
doi:10.1007/s00521-005-0005-y fatcat:p4h6brns3bdnddbn74jds5hksm

Logistic regression product-unit neural networks for mapping Ridolfia segetum infestations in sunflower crop using multitemporal remote sensed data

P.A. Gutiérrez, F. López-Granados, J.M. Peña-Barragán, M. Jurado-Expósito, C. Hervás-Martínez
2008 Computers and Electronics in Agriculture  
neural networks (EPUNNs), logistic regression (LR) and two different combinations of both (logistic regression using product units (LRPU) and logistic regression using initial covariates and product units  ...  In this paper, we used aerial imagery taken in mid-May, mid-June and mid-July according to different phenological stages of R. segetum and sunflower to evaluate the potential of evolutionary product-unit  ...  Evolutionary product-unit neural networks.  ... 
doi:10.1016/j.compag.2008.06.001 fatcat:ewygx5woqffolb6ywghux4ibgu

Evolutionary product unit based neural networks for regression

Alfonso Martínez-Estudillo, Francisco Martínez-Estudillo, César Hervás-Martínez, Nicolás García-Pedrajas
2006 Neural Networks  
This paper presents a new method for regression based on the evolution of a type of feed-forward neural networks whose basis function units are products of the inputs raised to real number power.  ...  Nevertheless, the training of product unit based networks poses several problems, since local learning algorithms are not suitable for these networks due to the existence of many local minima on the error  ...  This work has been supported in part by the Projects TIC2002-04036-C05-02 and TIN2005-08386-CO5-02 of the Spanish Comisión Interministerial de Ciencia y Tecnología and FEDER funds.  ... 
doi:10.1016/j.neunet.2005.11.001 pmid:16481148 fatcat:jdthqv5tanhhvfkzthd6wexv7e

Air Pollution Modelling by Machine Learning Methods

Petra Vidnerová, Roman Neruda
2021 Modelling  
All the methods considered achieved vital results, deep neural networks exhibited the best generalization ability, and regularization networks with product kernels achieved the best fitting of the training  ...  composite kernels, and deep neural networks.  ...  Author Contributions: Both authors contributed to the manuscript equally. All authors have read and agreed to the published version of the manuscript.  ... 
doi:10.3390/modelling2040035 fatcat:zniw775ih5fhzjs3ieqqdmwzlu

Memetic Pareto Evolutionary Artificial Neural Networks for the Determination of Growth Limits of Listeria Monocytogenes

J.C. Fernández, P.A. Gutiérrez, C. Hervás, F.J. Martínez
2008 2008 Eighth International Conference on Hybrid Intelligent Systems  
The main objective of this work is to automatically design neural network models with sigmoidal basis units for classification tasks, so that classifiers are obtained in the most balanced way possible  ...  We present a Memetic Pareto Evolutionary NSGA2 (MPENSGA2) approach based on the Pareto-NSGAII evolution (PNSGAII) algorithm.  ...  Specifically we investigate the generation of neural network classifiers based on two objectives: the correct classification rate, C, and the sensitivity, S.  ... 
doi:10.1109/his.2008.13 dblp:conf/his/FernandezGHM08 fatcat:mh7jiifxw5corjrijv6plbc63m

Permanent disability classification by combining evolutionary Generalized Radial Basis Function and logistic regression methods

A. Castaño, Francisco Fernández-Navarro, P.A. Gutiérrez, César Hervás-Martínez
2012 Expert systems with applications  
Our approach has been validated with a real problem of disability classification, to evaluate its effectiveness.  ...  Experimental results show that this approach is able to achieve good generalization performance.  ...  However, alternatives to MLP emerged in the last few years: Product Unit Neural Network (PUNN) models are an alternative to MLPs and are based on multiplicative neurons instead of additive ones.  ... 
doi:10.1016/j.eswa.2012.01.186 fatcat:tkij632alzforglrdyuirgk5di

The Criticality of Spare Parts Evaluating Model Using Artificial Neural Network Approach [chapter]

Lin Wang, Yurong Zeng, Jinlong Zhang, Wei Huang, Yukun Bao
2006 Lecture Notes in Computer Science  
This paper presents artificial neural networks (ANNs) for the criticality evaluating of spare parts in a power plant.  ...  The reliability of the models was tested by comparing their classification ability with a hold-out sample and an external data set.  ...  Artificial neural network (ANN) is an artificial intelligence based technique, which is applicable to the classification process.  ... 
doi:10.1007/11758501_97 fatcat:hhj2q7scnzb57bfp7jo74skrby
« Previous Showing results 1 — 15 out of 2,174 results