A modified artificial fish swarm algorithm for the optimization of extreme learning machines

Authors:
João Fausto Lorenzato de Oliveira;Teresa B. Ludermir
Affiliations:
Center of Informatics, Federal University of Pernambuco, Recife, PE, Brazil;Center of Informatics, Federal University of Pernambuco, Recife, PE, Brazil
Venue:
ICANN'12 Proceedings of the 22nd international conference on Artificial Neural Networks and Machine Learning - Volume Part II
Year:
2012

Citing 6
Cited 0

Neural Networks: A Comprehensive Foundation

Neural Networks: A Comprehensive Foundation
Differential Evolution – A Simple and Efficient Heuristic for Global Optimization over Continuous Spaces

Journal of Global Optimization
Fundamentals of Computational Swarm Intelligence

Fundamentals of Computational Swarm Intelligence
Rapid and brief communication: Evolutionary extreme learning machine

Pattern Recognition
ICGA-PSO-ELM Approach for Accurate Multiclass Cancer Classification Resulting in Reduced Gene Sets in Which Genes Encoding Secreted Proteins Are Highly Represented

IEEE/ACM Transactions on Computational Biology and Bioinformatics (TCBB)
Evolutionary extreme learning machine – based on particle swarm optimization

ISNN'06 Proceedings of the Third international conference on Advances in Neural Networks - Volume Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

Neural networks have been largely applied into many real world pattern classification problems. During the training phase, every neural network can suffer from generalization loss caused by overfitting, thereby the process of learning is highly biased. For this work we use Extreme Learning Machine which is an algorithm for training single hidden layer neural networks, and propose a novel swarm-based method for optimizing its weights and improving generalization performance. The algorithm presents the basic Artificial Fish Swarm Algorithm (AFSA) and some features from Differential Evolution (Crossover and Mutation) to improve the quality of the solutions during the search process. The results of the simulations demonstrated good generalization capacity from the best individuals obtained in the training phase.