Permutation Free Encoding Technique for Evolving Neural Networks

Authors:
Anupam Das;Md. Shohrab Hossain;Saeed Muhammad Abdullah;Rashed Ul Islam
Affiliations:
Department of Computer Science and Engineering, Bangladesh University of Engineering and Technology, Dhaka, Bangladesh;Department of Computer Science and Engineering, Bangladesh University of Engineering and Technology, Dhaka, Bangladesh;Department of Computer Science and Engineering, Bangladesh University of Engineering and Technology, Dhaka, Bangladesh;Department of Computer Science and Engineering, Bangladesh University of Engineering and Technology, Dhaka, Bangladesh
Venue:
ISNN '08 Proceedings of the 5th international symposium on Neural Networks: Advances in Neural Networks
Year:
2008

Citing 4
Cited 0

On the approximate realization of continuous mappings by neural networks

Neural Networks
A new evolutionary system for evolving artificial neural networks

IEEE Transactions on Neural Networks
Exploring constructive cascade networks

IEEE Transactions on Neural Networks
Use of a quasi-Newton method in a feedforward neural network construction algorithm

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a new evolutionary system using genetic algorithm for evolving artificial neural networks (ANNs). The proposed algorithm is "Permutation free Encoding Technique for Evolving Neural Networks"(PETENN) that uses a novel encoding scheme for representing ANNs. Existing genetic algorithms (GAs) for evolving ANNs suffer from the permutation problem, resulting from the recombination operator. Evolutionary Programming (EP) does not use recombination operator entirely. But the proposed encoding scheme avoids permutation problem by applying a sorting technique. PETENN uses two types of recombination operators that ensure automatic addition or deletion of nodes or links during the crossover process. The evolutionary system has been implemented and applied to a number of benchmark problems in machine learning and neural networks. The experimental results show that the system can dynamically evolve ANN architectures, showing competitiveness and, in some cases, superiority in performance.