Abstract
In this paper ensembles of self organizing NNs through fusion are introduced. In these ensembles not the output signals of the base learners are combined, but their architectures are properly merged. Merging algorithms for fusion and boosting-fusion-based ensembles of SOMs, GSOMs and NG networks are presented and positively evaluated on benchmarks from the UCI database. © 2009 Springer Berlin Heidelberg.
Cite
CITATION STYLE
Saavedra, C., Salas, R., Allende, H., & Moraga, C. (2009). Fusion of topology preserving neural networks. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5572 LNAI, pp. 517–524). https://doi.org/10.1007/978-3-642-02319-4_62
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
