SYMBIONT: A Cooperative Evolutionary Model for Evolving Artificial Neural Networks for Classification
Nicolás García‐Pedrajas, César Hervás‐Martínez, José Muñoz Pérez · Studies in fuzziness and soft computing · 2002
A new cooperative evolutionary model, called Symbiont , for evolving artificial neural networks is presented in this paper. This model is based on the idea of developing subnetworks, called nodules , that must cooperate to form a solution, instead of evolving a complete network. The performance of the model in solving two real-world problems of classification is compared with a multilayer perceptron trained using back-propagation. Symbiont has proved to show better generalization than the multilayer perceptron and to evolve smaller networks.