PubMed · 17131660
An optimization methodology for neural network weights and architectures.
Abstract
This paper introduces a methodology for neural network global optimization. The aim is the simultaneous optimization of multilayer perceptron (MLP) network weights and architectures, in order to generate topologies with few connections and high classification performance for any data sets. The approach combines the advantages of simulated annealing, tabu search and the backpropagation training algorithm in order to generate an automatic process for producing networks with high classification performance and low complexity. Experimental results obtained with four classification problems and one prediction problem has shown to be better than those obtained by the most commonly used optimization techniques.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Teresa B Ludermir, Akio Yamazaki, Cleber Zanchettin. 2006. An optimization methodology for neural network weights and architectures.. https://doi.org/10.1109/tnn.2006.881047. https://pubmed.health/record/pubmed%3A17131660/an-optimization-methodology-for-neural-network-weights-and-architectures
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite PubMed Health by Indiciae: BibTeX · RIS.