19 August 1993 Determination of optimal RBF network structure by canonical subspace analysis
Author Affiliations +
Abstract
An important consideration in designing an RBF network is the choice of the number of hidden units required for the network to generalize optimally. A new method, which is called canonical subspace analysis, is proposed for the selection of the number of hidden units. The numerical results show that with the number of the hidden units determined using the proposed method, minimum prediction errors are obtained.
© (1993) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Titus K. Y. Lo, Titus K. Y. Lo, John Litva, John Litva, Henry Leung, Henry Leung, } "Determination of optimal RBF network structure by canonical subspace analysis", Proc. SPIE 1966, Science of Artificial Neural Networks II, (19 August 1993); doi: 10.1117/12.152633; https://doi.org/10.1117/12.152633
PROCEEDINGS
10 PAGES


SHARE
Back to Top