Dynamical neural network architecture for continuous learning

MaRDI QIDQ2770461zbMATH OpenFDO

Authors Michael Tagscherer

Publication date 10 February 2002

zbMATH Keywords

algorithm ICEcontinuous learningneural network

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05)

Review text

An incremental construction algorithm ICE for continuous learning tasks for neural network is developed. The main advantages of the ICE-algorithm are: 1) the number of RBF-neurons and the number of local models of the hybrid network have not to be determined in advance (an important feature for fast initial learning); 2) the ability to simultaneously learn the target function and a confidence value for the network output. Finally a special version of the ICE-algorithm with asymmetric receptive fields is introduced. Similarities to fuzzy logic are intended.

Recommendations

• scientific article; zbMATH DE number 1304392
• The cascade neo-fuzzy neural network and its learning algorithm
• Incremental Structure Learning of Three-Layered Gaussian RBF Networks
• An incremental learning algorithm for function approximation
• scientific article; zbMATH DE number 1919441

Cited in

(7)

• A neural network architecture for incremental learning
• Dynamic updating of distributed neural representations using forward models
• scientific article; zbMATH DE number 1285309 (Why is no real title available?)
• scientific article; zbMATH DE number 1928643 (Why is no real title available?)
• Improving Generalization Capabilities of Dynamic Neural Networks
• Incremental learning with respect to new incoming input attributes
• Recurrent neural network architecture with pre-synaptic inhibition for incremental learning

This page was built for publication: Dynamical neural network architecture for continuous learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2770461)