Počet záznamů: 1
Hybrid Learning of RBF Networks
- 1.0404886 - UIVT-O 20020228 RIV CZ eng J - Článek v odborném periodiku
Neruda, Roman - Kudová, Petra
Hybrid Learning of RBF Networks.
Neural Network World. Roč. 12, č. 6 (2002), s. 573-585. ISSN 1210-0552
Grant CEP: GA ČR GA201/00/1489
Výzkumný záměr: AV0Z1030915
Klíčová slova: neural networks * RBF networks * gradient algorithm * three step learning * genetic algorithm
Kód oboru RIV: BA - Obecná matematika
Three different learning methods for RBF networks and their combinations are presented. Their performance is compared on two benchmarks problems: Two spirals and Iris plants. The results show that the three-step learning is usually the fastest, while the gradient learning achieves better precision.
Trvalý link: http://hdl.handle.net/11104/0003463
Počet záznamů: 1