Neural networks with asymmetric activation function for function approximation

Authors:
Gecynalda S. da S. Gomes;Teresa B. Ludermir;Leandro M. Almeida
Affiliations:
Centre of Informatics, The Federal University of Pernambuco, Recife, Pernambuco, Brazil;Centre of Informatics, The Federal University of Pernambuco, Recife, Pernambuco, Brazil;Centre of Informatics, The Federal University of Pernambuco, Recife, Pernambuco, Brazil
Venue:
IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Year:
2009

Citing 14
Cited 1

Multilayer feedforward networks are universal approximators

Neural Networks
Layered neural networks with Gaussian hidden units as universal approximations

Neural Computation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations

Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations
Approximation capabilities of multilayer feedforward networks

Neural Networks
Original Contribution: Multilayer feedforward networks with a nonpolynomial activation function can approximate any function

Neural Networks
Some new results on neural network approximation

Neural Networks
Lorentzian neural nets

Neural Networks
Neural Networks: A Comprehensive Foundation

Neural Networks: A Comprehensive Foundation
Effiicient BackProp

Neural Networks: Tricks of the Trade, this book is an outgrowth of a 1996 NIPS workshop
Sigmoidal Function Classes for Feedforward Artificial Neural Networks

Neural Processing Letters
Modelling and FDI of Dynamic Discrete Time Systems Using a MLP with a New Sigmoidal Activation Function

Journal of Intelligent and Robotic Systems
A max-piecewise-linear neural network for function approximation

Neurocomputing
Complementary Log-Log and Probit: Activation Functions Implemented in Artificial Neural Networks

HIS '08 Proceedings of the 2008 8th International Conference on Hybrid Intelligent Systems
Constructive feedforward neural networks using Hermite polynomial activation functions

IEEE Transactions on Neural Networks

Optimization of the weights and asymmetric activation function family of neural network for time series forecasting

Expert Systems with Applications: An International Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

The choice of activation functions may strongly influence complexity and performance of neural networks. However a limited number of activation functions have been used in practice for artificial neural networks. We propose the use of two new functions as asymmetric activation functions of neural networks and these defined functions are shown to satisfy the requirements of the universal approximation theorem.