Title :
A new compound structure of hierarchical neural networks
Author_Institution :
Dept. of Radio Eng., Beijing Univ. of Posts & Telecommun., China
Abstract :
A new compound structure of hierarchical neural networks is proposed in this paper. Compared with the present perceptron structure of hierarchical neural networks, this new type of neural networks is easier to train, and possesses good generalized ability. These advantageous properties have been proved in the experiment for speaker-independent speech recognition
Keywords :
neural nets; speech recognition; compound structure; hierarchical neural networks; speaker-independent speech recognition; Data processing; Fuzzy logic; Measurement standards; Multilayer perceptrons; Neural networks; Speech processing; Speech recognition; Testing; Vocabulary;
Conference_Titel :
Neural Networks Proceedings, 1998. IEEE World Congress on Computational Intelligence. The 1998 IEEE International Joint Conference on
Conference_Location :
Anchorage, AK
Print_ISBN :
0-7803-4859-1
DOI :
10.1109/IJCNN.1998.682306
