Optimisation of Artificial Neural Network Topology Applied in the Prosody Control in Text-to-Speech Synthesis

0404088 - UIVT-O 20000204 RIV DE eng C - Konferenční příspěvek (zahraniční konf.)
Šebesta, Václav - Tučková, Jana
Optimisation of Artificial Neural Network Topology Applied in the Prosody Control in Text-to-Speech Synthesis.
SOFSEM'2000: Theory and Practice of Informatics. Berlin: Springer, 2000 - (Hlaváč, V.; Jeffery, K.; Wiedermann, J.), s. 420-430. Lecture Notes in Computer Science, 1963. ISBN 3-540-41348-0. ISSN 0302-9743.
[SOFSEM 2000 /27./. Milovy (CZ), 25.11.2000-01.12.2000]
Grant CEP: GA AV ČR IAA2030801; GA ČR GV102/96/K087
Grant ostatní: COST(XE) 258
Výzkumný záměr: AV0Z1030915
Kód oboru RIV: BA - Obecná matematika

A possible way for the selection of significant input parameters, which are the most important ones from the point of view of influence on the output parameters, is presented. These parameters are later used for the training of ANN. We recognised that the approach based on mathematical logic, i.e. the GUHA method is also suitable for the determination of markers. Besides the minimisation of the number of elements in the input layer of ANN, also the number of neurons in hidden layers must be optimised.
Trvalý link: http://hdl.handle.net/11104/0124359